Currently we run the entirety of the Iceberg Java tests against Comet. Comet doesn't support accelerated ORC reads, and has no plans to do so. We should modify our Iceberg diffs to change the parameterized formats exercised to remove ORC, which we already fall back for. I would say we could skip table format V3 at the moment, but we'll hopefully start picking up support for V3 some features soon.
Currently we run the entirety of the Iceberg Java tests against Comet. Comet doesn't support accelerated ORC reads, and has no plans to do so. We should modify our Iceberg diffs to change the parameterized formats exercised to remove ORC, which we already fall back for. I would say we could skip table format V3 at the moment, but we'll hopefully start picking up support for V3 some features soon.