-1 (Binding)

I validated RC1 on macOS and Linux and found one packaging issue in the
source tarball, macOS metadata files are in the source tarball.


apache-xtable-0.5.0-incubating.src.tgz has 1,506 entries, and 753 of them
are macOS AppleDouble files (for example ._Foo.java). One is at the top
level, so extracting on Linux leaves ._apache-xtable-0.5.0-incubating next
to the source directory.


macOS tar folds them back into file attributes on extraction so folks
validating on Mac will miss this. To see them, extract with GNU tar on
Linux and run: find . -name '._*' | wc -l


The ones I sampled only hold the com.apple.provenance attribute. They don't
affect the build, tests or RAT, and the staged jars are clean but the 0.4.0
tarball doesn't have them and I would like to fix the script for generating
the tarball and create a second release candidate.


Environments

- macOS (Darwin 25.6, arm64), Java 11.0.30

- Ubuntu 26.04.1 (aarch64, clean Docker container with an empty Maven
repository), Java 11.0.32.1


Checked on both

[OK] Signatures and checksums

[OK] Release artifacts contain "apache" and "incubating" in the name

[OK] DISCLAIMER and LICENSE present and correct

[OK] NOTICE content and distribution year

[OK] Location and import of the KEYS file

[OK] README contains build instructions

[OK] No unexpected binary files in the source release

[OK] Apache RAT / license headers

[OK] Built from source with ./mvnw clean package, build cache disabled: 868
unit tests, 0 failures, 3 skipped

[OK] Version identifier format (0.5.0-incubating)

[OK] Staged Maven jars: signatures, checksums, structure, LICENSE and
NOTICE present


Git vs source release: apart from the ._* files above, the only differences
are the paths create_source_release.sh excludes and
dependency-reduced-pom.xml, which the build generates. No file contents
differ.


I ran a small integration test converting between Hudi, Delta, and Iceberg
and all are working as expected after upgrading the Hudi version.


Thanks,

Tim

On Sat, Nov 7, 2026 at 12:01 PM Vaibhav Kumar <[email protected]>
wrote:

> +1 (non-binding)
>
> macOS (Apple Silicon), OpenJDK 11.0.25 / 17.0.15, Maven wrapper 3.9.11
>
> [OK] Signature of the source release verifies against key
> 2F08D4856F4190C0C996920CF1901931503AB216 from the release KEYS
> [OK] SHA-512 checksum matches
> [OK] Artifact name contains "apache" and "incubating", version
> 0.5.0-incubating
> [OK] Tag release-0.5.0-incubating-rc1 is at
> 218cec874f0fefc08eadec7c1bd902813566e94d;
> the diff against the source release only shows the excludes in
> create_source_release.sh
> [OK] DISCLAIMER, LICENSE and NOTICE present (NOTICE 2024-2026)
> [OK] No binary files in the source release, ASF headers present,
> apache-rat:check clean (0 unapproved)
> [OK] Built from source with ./mvnw clean install -DskipTests on JDK 11
> and on JDK 17: BUILD SUCCESS on all modules
> [OK] Staged Maven repository (orgapachextable-1017) has the expected
> modules; xtable-utilities is not staged, as noted in the vote email
>
> Minor, non-blocking: with a system Maven newer than the wrapper (3.9.16),
> the build fails on JDK 11 at xtable-service, because quarkus-maven-plugin
> 3.2.12 declares Java 17 as a prerequisite and newer Maven enforces it. The
> README says to use Java 11, so it may be worth noting there to use ./mvnw
> (or JDK 17) for the service module.
>
> Thanks Tim for driving the release!
>
> On Sat, Nov 7, 2026 at 12:22 AM Vinish Reddy via dev <
> [email protected]>
> wrote:
>
> > +1 (binding)
> >
> > macOS (Apple Silicon), Temurin JDK 11.0.26, Maven wrapper, Spark 3.5.9
> >
> > [OK] Checked signature and checksum of the source release against the
> > release KEYS
> > [OK] Release artifacts contain "apache" and "incubating" in the name,
> > version 0.5.0-incubating
> > [OK] Diff between tag release-0.5.0-incubating-rc1 (218cec87) and the
> > source release matches, only the excludes in create_source_release.sh
> > differ
> > [OK] DISCLAIMER, LICENSE and NOTICE are present and correct (NOTICE
> > 2024-2026)
> > [OK] No unexpected binary files, ASF headers present, apache-rat:check
> > clean
> > [OK] Built from source with an empty local Maven repo (./mvnw clean
> package
> > -Dmaven.build.cache.enabled=false): BUILD SUCCESS on all 13 modules, 852
> > unit tests, 0 failures
> > [OK] Staged Maven artifacts (orgapachextable-1017): all 11 coordinates
> have
> > complete jar/sources/javadoc/pom sets (plus tests and bundled jars where
> > expected), 42 files in total. Each .asc verifies with gpg against Tim's
> key
> > from the release KEYS, and each .sha1 and .md5 matches the locally
> computed
> > digest.
> > [OK] The aws, hive-metastore, hudi-support-extensions and spark-runtime
> > bundles each carry one LICENSE, one NOTICE and the DISCLAIMER. Every text
> > under META-INF/licenses matches classes in the jar, and
> > validate_bundled_license_texts.py passes on the staged jars. #701 and
> #865
> > from the 0.4.0 vote are resolved. No Category X.
> > [OK] Sources jars carry no bundle license texts (#1000)
> >
> > Functional tests with the staged jars (Spark 3.5.9, Hudi 1.2.1, Iceberg
> > 1.9.2, Delta 3.3.2 with Delta Kernel 4.0.0), following the docs in #995:
> > - XTableSparkSync: Hudi table versions 6, 8 and 9 (written by Hudi
> 0.15.0,
> > 1.0.2 and 1.2.1) to Iceberg and Delta, row for row equal
> > - Iceberg to Hudi target, table version 6 (java engine with
> > hudi-java-client) and table version 9 (spark engine)
> > - Secondary index built with XTableSparkSync and with RunSync from the
> > utilities jar. With the SQL extensions from the staged runtime jar, an IN
> > lookup read 2 of 40 files and returned the same rows as without the
> index.
> > - Iceberg pluggable table format with the four staged jars: insert,
> upsert
> > and delete, Hudi and Iceberg reads match
> >
> > Non-blocking:
> > - The release notes in #1006 list only the release PRs. It would be good
> to
> > add the headline features (Hudi 1.x and table version 9, secondary index,
> > Iceberg pluggable table format).
> >
> > Thanks Tim for driving the release!
> >
> > Vinish
> >
> > On Thu, Nov 05, 2026 06:27 PM, Tim Brown <[email protected]> wrote:
> >
> > > Hi everyone,
> > >
> > > Please review and vote on the release candidate #1 for the version
> 0.5.0,
> > > as follows:
> > >
> > > [ ] +1, Approve the release
> > >
> > > [ ] -1, Do not approve the release (please provide specific comments)
> > >
> > > The complete staging area is available for your review, which includes:
> > >
> > > * GH release notes [1],
> > >
> > > * the official Apache source release to be deployed to dist.apache.org
> > > [2],
> > > which is signed with the key with fingerprint
> > > 2F08D4856F4190C0C996920CF1901931503AB216 [3],
> > >
> > > * all artifacts to be deployed to the Maven Central Repository [4],
> > >
> > > * source code tag "release-0.5.0-incubating-rc1" [5], at commit
> > > 218cec874f0fefc08eadec7c1bd902813566e94d.
> > >
> > > Note: the xtable-utilities bundled jar is not a release artifact and is
> > not
> > > deployed to Maven Central. It includes jol-core (GPL-2.0 with Classpath
> > > Exception).
> > >
> > > The vote will be open for at least 72 hours. It is adopted by majority
> > > approval, with at least 3 PMC affirmative votes.
> > >
> > > Thanks,
> > >
> > > Tim Brown
> > >
> > > [1] https://github.com/apache/incubator-xtable/issues/1006
> > >
> > > [2]
> > >
> > >
> >
> https://dist.apache.org/repos/dist/dev/incubator/xtable/0.5.0-incubating-rc1/
> > >
> > > [3] https://dist.apache.org/repos/dist/release/incubator/xtable/KEYS
> > >
> > > [4]
> > >
> https://repository.apache.org/content/repositories/orgapachextable-1017/
> > >
> > > [5]
> > >
> > >
> >
> https://github.com/apache/incubator-xtable/releases/tag/release-0.5.0-incubating-rc1
> > >
> >
>

Reply via email to