Re: [DISCUSS] Remove snapshot-id from IRC SetStatisticsUpdate

Hey Christian, Do you want to follow up with a PR? I was just reviewing a PR on PyIceberg and came to the same conclusion. Kind regards, Fokko Op di 17 dec 2024 om 23:36 schreef Marc Cenac : > +1 to removing this redundancy in the REST spec a

Re: [DISCUSS] Remove snapshot-id from IRC SetStatisticsUpdate

+1 to removing this redundancy in the REST spec and Java implementation On Tue, Dec 17, 2024 at 12:10 PM Kevin Liu wrote: > Hey Christian, > > Thanks for bringing this up! We also noticed this issue while implementing > table statistics in Python [1]. > I'm in favor of removing the outer field.

Re: [DISCUSS] Remove snapshot-id from IRC SetStatisticsUpdate

Hey Christian, Thanks for bringing this up! We also noticed this issue while implementing table statistics in Python [1]. I'm in favor of removing the outer field. Since this is part of the spec change, we would need to follow the proper deprecation and removal path, similar to what we did for `la

Re: [DISCUSS] Remove snapshot-id from IRC SetStatisticsUpdate

Hi, I saw the discussion on Slack. Yeah, it's redundant. I know some catalogs only consider the snapshot id in SetStatisticsUpdate. Regards JB On Fri, Dec 13, 2024 at 8:03 PM Christian wrote: > > Dear all, > > I believe we currently have a redundancy in the IRC SetStatisticsUpdate [1]. > SetSta

Re: [DISCUSS] Remove snapshot-id from IRC SetStatisticsUpdate

Hey Christian, Great catch, I would also be in favor of removing the outer one. I don't see any value in having them both. Kind regards, Fokko Op ma 16 dec 2024 om 14:26 schreef Jean-Baptiste Onofré : > Hi, > > I saw the discussion on Slack. Yeah, it's redundant. > I know some catalogs only con