+1 (binding) On Fri, Sep 4, 2026 at 6:55 PM Sung Yun <[email protected]> wrote:
> +1 (binding) > > Thanks for driving this Andrei. I agree that this is a good way to expose > semantic metadata in a catalog controlled way, on the grounds Ryan cited > above. > > Sung > > On 2026/09/04 00:29:57 Kevin Liu wrote: > > +1 (binding) > > > > Thanks for working on this. > > > > On Thu, Sep 3, 2026 at 1:53 PM Amogh Jahagirdar <[email protected]> > wrote: > > > > > +1 (binding) Thanks Andrei! > > > > > > On Thu, Sep 3, 2026 at 2:05 PM Anoop Johnson <[email protected]> wrote: > > > > > >> +1 (non-binding). Thanks for all the work on this, Andrei! > > >> > > >> On Thu, Sep 3, 2026 at 1:01 PM Ryan Blue <[email protected]> wrote: > > >> > > >>> +1 (binding) > > >>> > > >>> I think it is good to have a way to expose catalog metadata that does > > >>> not require overriding or reserving table properties. > > >>> > > >>> On Thu, Sep 3, 2026 at 8:58 AM Fokko Driesprong <[email protected]> > > >>> wrote: > > >>> > > >>>> +1 (binding) > > >>>> > > >>>> Thanks for driving this, Andrei! > > >>>> > > >>>> Kind regards, > > >>>> Fokko > > >>>> > > >>>> On 2026/09/03 15:55:55 Daniel Weeks wrote: > > >>>> > +1 (binding) > > >>>> > > > >>>> > Thanks, Andrei, for all the work on this! > > >>>> > > > >>>> > On Wed, Sep 2, 2026 at 2:48 PM Andrei Tserakhau via dev < > > >>>> > [email protected]> wrote: > > >>>> > > > >>>> > > Hi everyone, > > >>>> > > > > >>>> > > Following the discussion on IRC Labels and the catalog syncs, > I'd > > >>>> like to > > >>>> > > restart the vote specifically on the spec change: add optional > > >>>> > > catalog-provided labels on "LoadTableResult" and > "LoadViewResult". > > >>>> > > > > >>>> > > The proposal is intentionally limited to a read-only, flat > key-value > > >>>> > > representation of catalog-provided metadata. Structured tag > > >>>> entities and > > >>>> > > write APIs remain separate follow-ups. > > >>>> > > > > >>>> > > This vote is on the REST spec change in [3]. The other links > below > > >>>> are > > >>>> > > supporting context and implementations, and are not part of the > > >>>> vote. > > >>>> > > > > >>>> > > Discussion: > > >>>> > > [1] > > >>>> https://lists.apache.org/thread/0mpgn3p2xhlf5165m40oxbptmcf7f87m > > >>>> > > > > >>>> > > Proposal: > > >>>> > > [2] https://github.com/apache/iceberg/issues/15521 > > >>>> > > > > >>>> > > Spec PR (subject of this vote): > > >>>> > > [3] https://github.com/apache/iceberg/pull/15750 > > >>>> > > > > >>>> > > Design doc: > > >>>> > > [4] > > >>>> > > > > >>>> > https://docs.google.com/document/d/1aj-6JlfBiMYEEVtNuh5WLMOrRQiMCcyYUGbouPM4hXI/edit?usp=drivesdk > > >>>> > > > > >>>> > > Reference implementations / POCs: > > >>>> > > [5] Core read-path support: > > >>>> https://github.com/apache/iceberg/pull/17337 > > >>>> > > [6] Core "SupportsLabels": > > >>>> https://github.com/apache/iceberg/pull/17410 > > >>>> > > [7] Spark exposure: > https://github.com/apache/iceberg/pull/17411 > > >>>> > > [8] Gravitino IRC labels POC: > > >>>> > > https://github.com/laskoviymishka/gravitino/pull/1 > > >>>> > > > > >>>> > > Catalog sync recording: > > >>>> > > [9] https://www.youtube.com/watch?v=lCTV-Rhiwyc > > >>>> > > > > >>>> > > Thanks everyone for the discussion and feedback! > > >>>> > > > > >>>> > > Vote: > > >>>> > > > > >>>> > > [ ] +1 Adopt the spec change in [3] > > >>>> > > [ ] 0 No opinion > > >>>> > > [ ] -1 Do not adopt, because... > > >>>> > > > > >>>> > > This vote will be open for at least 72 hours. > > >>>> > > > > >>>> > > Thanks, > > >>>> > > Andrei > > >>>> > > > > >>>> > > > >>>> > > >>> > > >
