To make sure we don't lose track of this, https://github.com/apache/lucene/issues/13671 is a blocker for 10.0. There's a quick path forward here if we get into a time-crunch situation; we can revert some recent changes to drill-sideways (specifically DrillSideways and DrillSidewaysQuery) along with the new method that was added in IndexSearcher and ship—but then the newly-added sandbox faceting code (GH#13568) won't be able to work with drill-sideways.
As a more preferred solution, with the time we have before freeze I'd like to see if we can find a path forward that allows the new faceting code to work with drill-sideways while not adding API surface area to IndexSearcher. I've marked GH#13671 with the 10.0 milestone. Not sure if there's a better way to capture it as a blocker. Cheers, -Greg On Thu, Aug 22, 2024 at 6:06 AM Chris Hegarty <christopher.hega...@elastic.co.invalid> wrote: > > > > On 15 Aug 2024, at 10:52, Chris Hegarty <christopher.hega...@elastic.co> > wrote: > > > >> ... > >> > >> Chris, Uwe: I also wanted to check with you if this timeline works well > with regards to supporting Java 23 in 9.last and 10.0? > > > > Yes, this works for JDK 23. > > > > While JDK 23 is scheduled to ship on 17th Sep, there is already an > initial release candidate, which is all we need really need for testing and > verification. The Memory Segment code should be fine as-is, but the code > using the Panama Vector API will minimally need to be tested and the > runtime check bumped. I’ll get this done soon. > > Done. Unless I missed something. See > https://github.com/apache/lucene/pull/13678 > > -Chris > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org > For additional commands, e-mail: dev-h...@lucene.apache.org > >
