Hi folks, Here is the voting thread for the 1.12.0 RC1: https://lists.apache.org/thread/ptq6gyf6gcd4xwwkn2q3yrt9lxr7ojdk Please vote on the release when you get a chance. Thanks.
On Mon, Sep 21, 2026 at 10:09 AM Neelesh Salian <[email protected]> wrote: > Hi folks, > > The fix for https://github.com/apache/iceberg/issues/18191 has been > merged. > > And per the discussion, the behavior change for geo types will go ahead in > 1.12. The fallback PR has been closed: > https://github.com/apache/iceberg/pull/18169 > > We are cutting a new RC for this release. > I'll open a new thread for the voting on the new RC and link it here when > ready. > > > > On Fri, Sep 18, 2026 at 13:22 Neelesh Salian <[email protected]> > wrote: > >> Thanks for raising those concerns Steven. >> Looks like there is a PR: https://github.com/apache/iceberg/pull/18169 to >> fix it. If we want the behavior change or want to revert it, we should >> decide here. >> >> I'm awaiting the current voting to complete to see where we stand before >> we take a decision to proceed or cut a new Release candidate. >> >> >> On Fri, Sep 18, 2026 at 11:33 AM Steven Wu <[email protected]> wrote: >> >>> Sorry, I should have posted my comment in the discussion thread (not the >>> 1.12.0 RC0 voting thread). Pasting it here. >>> ------------------------------ >>> I want to call out a behavior change in 1.12. >>> >>> PR #16765 <https://github.com/apache/iceberg/pull/16765> changed >>> GeometryType / GeographyType so toString() always writes the resolved >>> defaults instead of omitting them. >>> >>> *Before*: Default instances collapsed to the bare type name. Geometry >>> with no CRS printed geometry. Geography printed geography, >>> geography(crs), or geography(crs, algorithm) depending on what was >>> stored as unset. >>> >>> *After*: the constructor stores omitted CRS / algorithm as the Iceberg >>> defaults, and toString() always includes them: >>> >>> - geometry → geometry(OGC:CRS84) >>> - geography → geography(OGC:CRS84, spherical) >>> >>> *Implications* >>> >>> - Read path is backward compatible, as SchemaParser or >>> Types.fromTypeName can parse the type string to the same Type object >>> - Write path is not string stable. >>> >>> *Question*: Do we consider the type's `toString()` change a breaking >>> change? Personally, I am okay with 1.12 moving forward with the changed >>> behavior since read is compatible. But I wanted to bring this to the >>> community's attention. >>> >>> On Thu, Sep 17, 2026 at 5:30 PM Neelesh Salian <[email protected]> >>> wrote: >>> >>>> Hi folks, >>>> >>>> The RC0 has been created and the vote thread can be found here: >>>> https://lists.apache.org/thread/7qtxrysl1w5sjs7o74y31snt0jrkf5qb >>>> Thank you for all the work. >>>> >>>> On Thu, Sep 17, 2026 at 9:38 AM Neelesh Salian < >>>> [email protected]> wrote: >>>> >>>>> Hi Alex, >>>>> >>>>> Thanks for the note. >>>>> It looks like https://github.com/apache/iceberg/pull/17713 has >>>>> already been merged, so it should be in the release. >>>>> I would defer the https://github.com/apache/iceberg/pull/13979 over >>>>> to the next release given that it isn't necessarily a blocker for 1.12. >>>>> I've gone ahead and added it to the 1.13.0. milestone for now. >>>>> >>>>> >>>>> On Thu, Sep 17, 2026 at 9:00 AM Alex Reid <[email protected]> >>>>> wrote: >>>>> >>>>>> Hi All, >>>>>> >>>>>> Is it possible to get the following PR into the 1.12 release? >>>>>> https://github.com/apache/iceberg/pull/17713 -- This is another fix >>>>>> for Kafka-Connect to prevent committing already committed files in >>>>>> certain >>>>>> rebalance cases. >>>>>> >>>>>> Also, I'd like to discuss the "referenced-by" implementation - >>>>>> https://github.com/apache/iceberg/pull/13979 >>>>>> This was added to the spec last year ( >>>>>> https://github.com/apache/iceberg/pull/13810), but the actual REST >>>>>> client implementation, etc. Is it possible to try and get this into the >>>>>> 1.12 release as well? >>>>>> >>>>>> Thanks, >>>>>> Alex >>>>>> >>>>>> On Wed, Sep 16, 2026 at 1:58 PM Neelesh Salian < >>>>>> [email protected]> wrote: >>>>>> >>>>>>> Thanks for the note Gabor. >>>>>>> I believe we're now in a good place with the >>>>>>> https://github.com/apache/iceberg/pull/17984 fix and it has been >>>>>>> merged. >>>>>>> I do think the interface change needs more debate and I don't see it >>>>>>> needing to be rushed for this release. >>>>>>> >>>>>>> Separately, we are ready to cut the RC. Thanks again for all the >>>>>>> work across the community. I followed up with the voting. >>>>>>> >>>>>>> On Sun, Sep 13, 2026 at 11:41 PM Gábor Kaszab < >>>>>>> [email protected]> wrote: >>>>>>> >>>>>>>> Hi All, >>>>>>>> >>>>>>>> Thanks for the update, Neelesh! >>>>>>>> Mentioning *encryption keys* reminds me that I bumped into a PR >>>>>>>> <https://github.com/apache/iceberg/pull/16353> the other day >>>>>>>> saying that currently there is *no mechanism to clean up* unused >>>>>>>> encryption keys from TableMetadata. I'm wondering how severe of an >>>>>>>> issue this is, should we add it into this release or can it wait. >>>>>>>> >>>>>>>> Thanks, >>>>>>>> Gabor >>>>>>>> >>>>>>>> Neelesh Salian <[email protected]> ezt írta (időpont: 2026. >>>>>>>> szept. 12., Szo, 3:44): >>>>>>>> >>>>>>>>> Hi folks, >>>>>>>>> >>>>>>>>> Thank you for all the great work across the community for getting >>>>>>>>> the fixes in. >>>>>>>>> All of the previous correctness fixes have been merged. >>>>>>>>> >>>>>>>>> The only one that prevails: >>>>>>>>> [#17984 <https://github.com/apache/iceberg/pull/17984>] - Core, >>>>>>>>> Hive: Commit manifest list encryption keys with the snapshot that >>>>>>>>> uses them >>>>>>>>> Once this merges, the release will be cut and I'll follow up with >>>>>>>>> the formalities. >>>>>>>>> >>>>>>>>> As always, please reach out if you have questions. >>>>>>>>> >>>>>>>>> Thank you. >>>>>>>>> >>>>>>>>> On Tue, Sep 1, 2026 at 10:35 PM Neelesh Salian < >>>>>>>>> [email protected]> wrote: >>>>>>>>> >>>>>>>>>> Hi folks, >>>>>>>>>> >>>>>>>>>> Thank you to everyone who has been reviewing and merging 1.12.0 >>>>>>>>>> work over the >>>>>>>>>> past few weeks. We are close. The milestone is down to a few open >>>>>>>>>> PRs, and I would >>>>>>>>>> like to get through them so we can cut RC0 in the next few days. >>>>>>>>>> Given the long weekend in the US, I'll be sure to keep the vote >>>>>>>>>> open for longer to give enough time. >>>>>>>>>> >>>>>>>>>> If you have some time, reviews on the following would help most. >>>>>>>>>> I have grouped them by where they stand. >>>>>>>>>> >>>>>>>>>> *Approved, needs more reviews + merge:* >>>>>>>>>> - #17552 <https://github.com/apache/iceberg/pull/17552> Kafka >>>>>>>>>> Connect: coordinator only commits offsets when greater than existing >>>>>>>>>> offsets >>>>>>>>>> - #15989 <https://github.com/apache/iceberg/pull/15989> Core, >>>>>>>>>> REST: Fix path segment encoding to use RFC 3986 percent-encoding >>>>>>>>>> *Need review:* >>>>>>>>>> - #17774 <https://github.com/apache/iceberg/pull/17774> Core: >>>>>>>>>> Interleave Z-order bits a byte at a time with a lookup table >>>>>>>>>> - #17669 <https://github.com/apache/iceberg/pull/17669> Spark: >>>>>>>>>> Fix Z-order NPE on null booleans and case-insensitive column >>>>>>>>>> resolution >>>>>>>>>> - #17557 <https://github.com/apache/iceberg/pull/17557> Core, >>>>>>>>>> Parquet: Fix null counting for Parquet that does not have null_count >>>>>>>>>> - #17431 <https://github.com/apache/iceberg/pull/17431> Parquet: >>>>>>>>>> Guard ordered predicates on variant/nested types in the metrics row >>>>>>>>>> group >>>>>>>>>> filter >>>>>>>>>> - #17320 <https://github.com/apache/iceberg/pull/17320> Parquet: >>>>>>>>>> Fix nested initial default applied when ancestor struct is null >>>>>>>>>> - #16507 <https://github.com/apache/iceberg/pull/16507> API, >>>>>>>>>> Core: Add exceptions for OAuth2 token endpoint errors >>>>>>>>>> *Needs another round (changes requested):* >>>>>>>>>> - #17764 <https://github.com/apache/iceberg/pull/17764> Core: >>>>>>>>>> Fix commit validation when a data file has multiple DVs across >>>>>>>>>> snapshots >>>>>>>>>> >>>>>>>>>> If any of these should move to 1.12.1 rather than block the >>>>>>>>>> release, please say >>>>>>>>>> so on the PR or here and I will take it off the milestone. >>>>>>>>>> Otherwise the plan is to cut RC0 once they are resolved. >>>>>>>>>> >>>>>>>>>> The Full Milestone is here: >>>>>>>>>> https://github.com/apache/iceberg/milestone/60 >>>>>>>>>> >>>>>>>>>> Thanks again for helping get us here. >>>>>>>>>> >>>>>>>>>> Neelesh. >>>>>>>>>> >>>>>>>>>> On Thu, Aug 27, 2026 at 10:12 AM Neelesh Salian < >>>>>>>>>> [email protected]> wrote: >>>>>>>>>> >>>>>>>>>>> Hi folks, >>>>>>>>>>> >>>>>>>>>>> As mentioned during the community sync yesterday, I've cleaned >>>>>>>>>>> up the milestone to reflect what is needed for the 1.12 release. >>>>>>>>>>> There are a few new additions: >>>>>>>>>>> >>>>>>>>>>> Core: >>>>>>>>>>> - Tolerate multiple cross-snapshot DVs for the same data file in >>>>>>>>>>> commit validation: https://github.com/apache/iceberg/pull/17764 >>>>>>>>>>> >>>>>>>>>>> Updating the Flink version support: >>>>>>>>>>> - Flink: Upgrade to latest supported Flink versions 2.2 and 2.3 >>>>>>>>>>> without dropping 2.1 ( >>>>>>>>>>> https://github.com/apache/iceberg/pull/17849) >>>>>>>>>>> - Flink: Upgrade to latest supported Flink versions 2.2 and 2.3 >>>>>>>>>>> (https://github.com/apache/iceberg/pull/17845) >>>>>>>>>>> >>>>>>>>>>> The previous blocker: #16543 >>>>>>>>>>> <https://github.com/apache/iceberg/pull/16543> - Build: Clean >>>>>>>>>>> up shaded-jar LICENSE/NOTICE and add missing third-party notices >>>>>>>>>>> has been >>>>>>>>>>> merged. >>>>>>>>>>> >>>>>>>>>>> I'll communicate on this thread when we are past the active work >>>>>>>>>>> and when we are ready to cut an RC. >>>>>>>>>>> >>>>>>>>>>> Thank you again for all the contributions, reviews. >>>>>>>>>>> As always, please reach out if you have questions. >>>>>>>>>>> >>>>>>>>>>> Thank you. >>>>>>>>>>> >>>>>>>>>>> On Mon, Aug 24, 2026 at 2:31 PM Neelesh Salian < >>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>> >>>>>>>>>>>> Hi folks, >>>>>>>>>>>> >>>>>>>>>>>> Surfacing the following to close out before the RC is cut. >>>>>>>>>>>> >>>>>>>>>>>> *Blockers for the release:* >>>>>>>>>>>> >>>>>>>>>>>> - #16543 <https://github.com/apache/iceberg/pull/16543> - >>>>>>>>>>>> Build: Clean up shaded-jar LICENSE/NOTICE and add missing >>>>>>>>>>>> third-party >>>>>>>>>>>> notices. Follow-up to the LICENSE/NOTICE findings from the >>>>>>>>>>>> 1.11.0 RC4 vote. >>>>>>>>>>>> It has no approvals yet. IMO, we should not cut RC0 with this >>>>>>>>>>>> open. >>>>>>>>>>>> - #17552 <https://github.com/apache/iceberg/pull/17552> - >>>>>>>>>>>> Kafka Connect: coordinator only commits offsets when greater >>>>>>>>>>>> than existing >>>>>>>>>>>> offsets. When it merges it closes: #17551 >>>>>>>>>>>> <https://github.com/apache/iceberg/issues/17551> - data >>>>>>>>>>>> loss with default configs due to persistent zombie coordinators. >>>>>>>>>>>> - Related to the above Kafka Connect one #17340 >>>>>>>>>>>> <https://github.com/apache/iceberg/issues/17340>- duplicate >>>>>>>>>>>> data file commits after task restart - is a separate open issue >>>>>>>>>>>> on the >>>>>>>>>>>> milestone and still needs a fix >>>>>>>>>>>> >>>>>>>>>>>> *Decisions (needed before the RC0 cut):* >>>>>>>>>>>> >>>>>>>>>>>> - #17706 <https://github.com/apache/iceberg/pull/17706> - >>>>>>>>>>>> Remove writing of position delete files with row data (PDWR). >>>>>>>>>>>> Should this >>>>>>>>>>>> be in or out for 1.12.0. Please weigh in on the discussion >>>>>>>>>>>> thread: Dev >>>>>>>>>>>> List Thread for PDWR >>>>>>>>>>>> >>>>>>>>>>>> <https://lists.apache.org/thread/3tz63wok59hhqphswovvl9z0yrh0z7pl> >>>>>>>>>>>> >>>>>>>>>>>> *Correctness Fixes:* >>>>>>>>>>>> >>>>>>>>>>>> - #17669 <https://github.com/apache/iceberg/pull/17669> - >>>>>>>>>>>> Spark: Fix Z-order NPE on null booleans and case-insensitive >>>>>>>>>>>> column >>>>>>>>>>>> resolution >>>>>>>>>>>> - #17557 <https://github.com/apache/iceberg/pull/17557>- >>>>>>>>>>>> Core, Parquet: Fix null counting for parquet without null_count >>>>>>>>>>>> - #17431 <https://github.com/apache/iceberg/pull/17431> - >>>>>>>>>>>> Parquet: Guard ordered predicates on variant/nested types in >>>>>>>>>>>> metrics row >>>>>>>>>>>> group filter >>>>>>>>>>>> - #17320 <https://github.com/apache/iceberg/pull/17320> - >>>>>>>>>>>> Parquet: Fix nested initial default applied when ancestor >>>>>>>>>>>> struct is null. >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> re the Flink Docs: Cheng, thanks for the note. Would you be >>>>>>>>>>>> able to put up a PR to update, we can iterate on it? I wouldn't >>>>>>>>>>>> put it as a >>>>>>>>>>>> blocker for now, but worth getting the docs in a good place. >>>>>>>>>>>> >>>>>>>>>>>> If folks could review and help these in, it would be helpful >>>>>>>>>>>> before the RC. Partnering with Amogh to do the artifact >>>>>>>>>>>> generation, etc. >>>>>>>>>>>> Full milestone: https://github.com/apache/iceberg/milestone/60 >>>>>>>>>>>> >>>>>>>>>>>> I would not add anything additional to the milestone unless it >>>>>>>>>>>> is a blocker / bug or license issue going forward from today >>>>>>>>>>>> (08/24/26) so >>>>>>>>>>>> our release timeline is met. >>>>>>>>>>>> As always, please reach out if you have questions. >>>>>>>>>>>> >>>>>>>>>>>> Thank you. >>>>>>>>>>>> >>>>>>>>>>>> On Thu, Aug 20, 2026 at 11:46 PM Cheng Pan <[email protected]> >>>>>>>>>>>> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> Have a glance at the docs site, I found that the latest docs >>>>>>>>>>>>> still have many references to the old versions of engines >>>>>>>>>>>>> (especially >>>>>>>>>>>>> Flink, and Spark looks better) that Iceberg has already removed >>>>>>>>>>>>> support >>>>>>>>>>>>> for. Just wondering if someone familiar with Flink has a chance >>>>>>>>>>>>> to update >>>>>>>>>>>>> that. >>>>>>>>>>>>> >>>>>>>>>>>>> Thanks, >>>>>>>>>>>>> Cheng Pan >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On Aug 21, 2026, at 09:40, Manu Zhang <[email protected]> >>>>>>>>>>>>> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>> Hi all, >>>>>>>>>>>>> >>>>>>>>>>>>> For Spark 4.2.0 support, Szehon has a nice suggestion: merge >>>>>>>>>>>>> the source code while excluding it from release. In the latest >>>>>>>>>>>>> commit, I've >>>>>>>>>>>>> excluded it from both binary and source tarballs, and I believe >>>>>>>>>>>>> the PR is >>>>>>>>>>>>> ready to merge now (Thanks to all the reviewers!) >>>>>>>>>>>>> >>>>>>>>>>>>> Neelesh, >>>>>>>>>>>>> Can we also include #16028 >>>>>>>>>>>>> <https://github.com/apache/iceberg/pull/16028>, to mark >>>>>>>>>>>>> overrides of deprecated Spark API as deprecated? It's also close >>>>>>>>>>>>> to merging. >>>>>>>>>>>>> >>>>>>>>>>>>> On Wed, Aug 19, 2026 at 1:27 AM Neelesh Salian < >>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> Hi, >>>>>>>>>>>>>> >>>>>>>>>>>>>> Thanks for the comments Anurag, Anton and Alex. >>>>>>>>>>>>>> >>>>>>>>>>>>>> re Hilbert Curve PR: It has been added and is on track for >>>>>>>>>>>>>> the release. >>>>>>>>>>>>>> re Spark 4.2: This makes sense. The PR does touch a large >>>>>>>>>>>>>> surface and I was hoping we could get a baseline 4.2 support in >>>>>>>>>>>>>> and add >>>>>>>>>>>>>> more things along. I've removed it from the milestone for 1.12. >>>>>>>>>>>>>> re the REST changes: Alex, I've included 17709 and 17627 >>>>>>>>>>>>>> onto the milestone. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Thanks. >>>>>>>>>>>>>> >>>>>>>>>>>>>> On Tue, Aug 18, 2026 at 6:05 AM Alexandre Dutra < >>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> Hi all, >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> I'd like to request that #17709 be included in 1.12 as well. >>>>>>>>>>>>>>> This PR >>>>>>>>>>>>>>> implements the REST spec changes related to remote signing >>>>>>>>>>>>>>> configuration, which were introduced by #16822. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Without #17709, we would be shipping a spec change that is >>>>>>>>>>>>>>> not >>>>>>>>>>>>>>> implemented in any language. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Thanks, >>>>>>>>>>>>>>> Alex >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> [16822]: https://github.com/apache/iceberg/pull/16822 >>>>>>>>>>>>>>> [17627]: https://github.com/apache/iceberg/pull/17627 >>>>>>>>>>>>>>> [17709]: https://github.com/apache/iceberg/pull/17709 >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> On Tue, Aug 18, 2026 at 2:01 PM Anton Okolnychyi < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> > >>>>>>>>>>>>>>> > Btw, I would advise against shipping Spark 4.2 support. >>>>>>>>>>>>>>> Each new Spark release is a unique opportunity to address some >>>>>>>>>>>>>>> tech debt >>>>>>>>>>>>>>> because the DSv2 APIs are evolving. 4.2 is no different. I'd >>>>>>>>>>>>>>> suggest >>>>>>>>>>>>>>> getting the module into main to unblock development but skip it >>>>>>>>>>>>>>> from the >>>>>>>>>>>>>>> next Iceberg release. We can do another Iceberg release when >>>>>>>>>>>>>>> the Spark 4.2 >>>>>>>>>>>>>>> integration matures. >>>>>>>>>>>>>>> > >>>>>>>>>>>>>>> > пн, 17 серп. 2026 р. о 21:58 Anurag Mantripragada < >>>>>>>>>>>>>>> [email protected]> пише: >>>>>>>>>>>>>>> >> >>>>>>>>>>>>>>> >> Hi Neelesh, >>>>>>>>>>>>>>> >> >>>>>>>>>>>>>>> >> Thanks for tracking these items. Would you mind including >>>>>>>>>>>>>>> the Hilbert Curve clustering PR in this release? It is already >>>>>>>>>>>>>>> approved by >>>>>>>>>>>>>>> Russell. It will be a useful clustering strategy. >>>>>>>>>>>>>>> >> >>>>>>>>>>>>>>> >> ~ Anurag >>>>>>>>>>>>>>> >> >>>>>>>>>>>>>>> >> On Sun, Aug 16, 2026 at 6:56 PM Neelesh Salian < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Hi all, >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> We are targeting a 1.12.0 RC on or shortly after Aug 26. >>>>>>>>>>>>>>> The following PRs are still open on the 1.12.0 milestone and >>>>>>>>>>>>>>> need reviews >>>>>>>>>>>>>>> to make the release. >>>>>>>>>>>>>>> >>> If you own or can review any of these, please take a >>>>>>>>>>>>>>> pass this week. >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Blockers (need to land before the RC is cut): >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> #16543 Build: Clean up shaded-jar LICENSE/NOTICE and >>>>>>>>>>>>>>> add missing third-party notices >>>>>>>>>>>>>>> >>> #16449 All: Remove deprecated methods for 1.12.0 >>>>>>>>>>>>>>> >>> #17638 Core: Reduce visibility on scan planning response >>>>>>>>>>>>>>> builder for specsById and deleteFiles >>>>>>>>>>>>>>> >>> #17627 AWS: Remove deprecated class and require >>>>>>>>>>>>>>> SIGNER_ENDPOINT >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Correctness (great to get these in): >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> #17557 Core, Parquet: Fix null counting for parquet that >>>>>>>>>>>>>>> doesn't have null_count >>>>>>>>>>>>>>> >>> #15989 Core, REST: Fix path segment encoding to use RFC >>>>>>>>>>>>>>> 3986 percent-encoding >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Feature/enhancement candidates (in if ready, otherwise >>>>>>>>>>>>>>> they can be moved to 1.13): >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> #14984 Spark: Add support for 4.2.0 >>>>>>>>>>>>>>> >>> #16750 Spark 4.1: Implement SupportsReportOrdering DSv2 >>>>>>>>>>>>>>> API >>>>>>>>>>>>>>> >>> #16507 API, Core: Add exceptions for OAuth2 token >>>>>>>>>>>>>>> endpoint errors >>>>>>>>>>>>>>> >>> #16305 Core, Spark 4.1: Add K-way merge rewrite strategy >>>>>>>>>>>>>>> for pre-sorted data files >>>>>>>>>>>>>>> >>> #17520 Data, Kafka Connect: Enable Parquet variant >>>>>>>>>>>>>>> shredding for Kafka Connect and generic Record writes >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Full milestone: >>>>>>>>>>>>>>> https://github.com/apache/iceberg/milestone/60 >>>>>>>>>>>>>>> >>> If you have a feature(s), fix(es), or breaking change(s) >>>>>>>>>>>>>>> that are missed, please let me know so we can track them. >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Thank you for all the great work across the community. >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> Thanks, >>>>>>>>>>>>>>> >>> Neelesh >>>>>>>>>>>>>>> >>> >>>>>>>>>>>>>>> >>> On Thu, Aug 13, 2026 at 2:27 AM Péter Váry < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>> >>>>>>>>>>>>>>> >>>> Yeah, >>>>>>>>>>>>>>> >>>> I think we definitely want to decide and move forward >>>>>>>>>>>>>>> with the Flink version upgrade for the next Iceberg release. >>>>>>>>>>>>>>> >>>> >>>>>>>>>>>>>>> >>>> Felix Perez Diener via dev <[email protected]> >>>>>>>>>>>>>>> ezt írta (időpont: 2026. aug. 10., H, 15:47): >>>>>>>>>>>>>>> >>>>> >>>>>>>>>>>>>>> >>>>> Hello all, >>>>>>>>>>>>>>> >>>>> >>>>>>>>>>>>>>> >>>>> I see that support for Flink 2.3 was discussed for >>>>>>>>>>>>>>> release in Iceberg 1.12.0. Is that still the case? Would really >>>>>>>>>>>>>>> love to >>>>>>>>>>>>>>> have Flink 2.3 supported, since at Stripe we have started >>>>>>>>>>>>>>> upgrading to >>>>>>>>>>>>>>> Flink 2.3, but missing Iceberg support may make things >>>>>>>>>>>>>>> difficult. >>>>>>>>>>>>>>> >>>>> >>>>>>>>>>>>>>> >>>>> Best, >>>>>>>>>>>>>>> >>>>> Felix >>>>>>>>>>>>>>> >>>>> >>>>>>>>>>>>>>> >>>>> On Mon, Aug 10, 2026 at 7:14 AM Maximilian Michels < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >>>>>>>>>>>>>>> >>>>>> Thanks Neelesh for getting us started on the 1.12.0 >>>>>>>>>>>>>>> release! >>>>>>>>>>>>>>> >>>>>> >>>>>>>>>>>>>>> >>>>>> -Max >>>>>>>>>>>>>>> >>>>>> >>>>>>>>>>>>>>> >>>>>> On Sat, Aug 8, 2026 at 7:14 PM Neelesh Salian < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> > >>>>>>>>>>>>>>> >>>>>> > Thanks, Alex. Looks like the PR was added to the >>>>>>>>>>>>>>> milestone. >>>>>>>>>>>>>>> >>>>>> > >>>>>>>>>>>>>>> >>>>>> > Cheng, thank you for the note. I do agree that at >>>>>>>>>>>>>>> some point the default should move forward, but there are still >>>>>>>>>>>>>>> gaps to be >>>>>>>>>>>>>>> closed on Variant as well as Geo types. >>>>>>>>>>>>>>> >>>>>> > I don't think that would happen within the timeline >>>>>>>>>>>>>>> of 1.12. >>>>>>>>>>>>>>> >>>>>> > Let me look at the current state and see what is >>>>>>>>>>>>>>> pending, as well as improve the tracking. I'll start a separate >>>>>>>>>>>>>>> thread for >>>>>>>>>>>>>>> that effort. >>>>>>>>>>>>>>> >>>>>> > >>>>>>>>>>>>>>> >>>>>> > On Sat, Aug 8, 2026 at 8:19 AM Cheng Pan < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> I assume the Iceberg Java already includes a full >>>>>>>>>>>>>>> implementation of the Table Spec V3 features. Does the >>>>>>>>>>>>>>> community have plans >>>>>>>>>>>>>>> to switch the default table version from 2 to 3 for Iceberg >>>>>>>>>>>>>>> tables created >>>>>>>>>>>>>>> by Spark in the next release? This would help drive the >>>>>>>>>>>>>>> adoption of new >>>>>>>>>>>>>>> features such as Variant. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Thanks, >>>>>>>>>>>>>>> >>>>>> >> Cheng Pan >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> On Aug 6, 2026, at 16:18, Alexandre Dutra < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Hi Neelesh, >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> If possible, I'd like to have this PR included: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> https://github.com/apache/iceberg/pull/15989 >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> It's the fix for the path segment encoding issue >>>>>>>>>>>>>>> in Iceberg REST. The >>>>>>>>>>>>>>> >>>>>> >> PR got one review already but isn't approved yet. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Thanks, >>>>>>>>>>>>>>> >>>>>> >> Alex >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> On Wed, Aug 5, 2026 at 10:37 PM Oğuzhan Ünlü < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Hi Neelesh, >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> I'd like to get >>>>>>>>>>>>>>> https://github.com/apache/iceberg/pull/16507 included in >>>>>>>>>>>>>>> the milestone if possible. I previously asked for a review but >>>>>>>>>>>>>>> didn't hear >>>>>>>>>>>>>>> back. Not sure if that went to spam for some reason. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Thanks, >>>>>>>>>>>>>>> >>>>>> >> Oguzhan >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> On Wed, Aug 5, 2026 at 11:03 PM Neelesh Salian < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Hi folks, >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> As mentioned during the community sync this >>>>>>>>>>>>>>> morning, the release is on track. >>>>>>>>>>>>>>> >>>>>> >> A few updates: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> I plan to cut the branch / RC on or after August >>>>>>>>>>>>>>> 26, 2026 (after the community sync that morning). I'll be >>>>>>>>>>>>>>> working with >>>>>>>>>>>>>>> Amogh to help with the branching + artifacts. >>>>>>>>>>>>>>> >>>>>> >> There are still active items needing reviews. >>>>>>>>>>>>>>> Please take a look at the 1.12.0 milestone. >>>>>>>>>>>>>>> >>>>>> >> Calling out the Spark 4.2.0 PR: >>>>>>>>>>>>>>> https://github.com/apache/iceberg/pull/14984 - it needs >>>>>>>>>>>>>>> more eyes and review time to make this release. >>>>>>>>>>>>>>> >>>>>> >> I'll keep watching for blockers and bugs that >>>>>>>>>>>>>>> should be part of the release and surface them as needed, >>>>>>>>>>>>>>> including any >>>>>>>>>>>>>>> forward-porting to Spark 4.2. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> If you have a feature(s), fix(es), or breaking >>>>>>>>>>>>>>> change(s) you want considered for 1.12.0, please let me know. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Thanks, >>>>>>>>>>>>>>> >>>>>> >> Neelesh >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> On Fri, Jun 5, 2026 at 3:54 PM Steven Wu < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> If we develop good habit of regular releases, we >>>>>>>>>>>>>>> normally won't need to hold the release train to get things in >>>>>>>>>>>>>>> (except for >>>>>>>>>>>>>>> critical security and correctness issues) as the next one is >>>>>>>>>>>>>>> 2-3 months >>>>>>>>>>>>>>> away. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> On Fri, Jun 5, 2026 at 3:42 PM Neelesh Salian < >>>>>>>>>>>>>>> [email protected]> wrote: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Hi folks, >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> With 1.11.0 out the door, I want to open up the >>>>>>>>>>>>>>> planning around 1.12.0. >>>>>>>>>>>>>>> >>>>>> >> As Steven noted on the dev list (thread for 1.11), >>>>>>>>>>>>>>> 1.11.0 took roughly 8 months from 1.10.0. This is longer than >>>>>>>>>>>>>>> the 3-4 month >>>>>>>>>>>>>>> cadence we have previously discussed. >>>>>>>>>>>>>>> >>>>>> >> For 1.12.0 I'd like us to aim closer to that >>>>>>>>>>>>>>> target. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> I'll be acting as the release manager for this >>>>>>>>>>>>>>> release. Here are a few notes to kick things off: >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Tentative release target: late August / early >>>>>>>>>>>>>>> September 2026 (~3 months from 1.11.0). >>>>>>>>>>>>>>> >>>>>> >> Curation: We'll use the June 24 community sync as >>>>>>>>>>>>>>> the formal kickoff to surface in-flight work and call out >>>>>>>>>>>>>>> anything that >>>>>>>>>>>>>>> should land in this release. >>>>>>>>>>>>>>> >>>>>> >> Cadence reminder: If something doesn't make it in >>>>>>>>>>>>>>> 1.12, the next one is ~3 months out (1.13). >>>>>>>>>>>>>>> >>>>>> >> Mid-cycle check-in: I'll send a >>>>>>>>>>>>>>> state-of-the-release update around mid-July with what is in, >>>>>>>>>>>>>>> what is >>>>>>>>>>>>>>> in-flight, and what remains a blocker. >>>>>>>>>>>>>>> >>>>>> >> Tracking: The open issues and PRs targeting this >>>>>>>>>>>>>>> release are tagged with the Iceberg 1.12.0 milestone. I'll keep >>>>>>>>>>>>>>> this >>>>>>>>>>>>>>> milestone updated as things land in the repo. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> If you have a feature, fix, or breaking change you >>>>>>>>>>>>>>> want considered for 1.12.0, please reply on this thread or flag >>>>>>>>>>>>>>> it on the >>>>>>>>>>>>>>> June 24 sync. >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> Thanks, >>>>>>>>>>>>>>> >>>>>> >> Neelesh >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>> >> >>>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>
