Parquet tuning has always been like that, ever since it first came out in 2013. I worry with Iceberg that people think it's just a case of "use an Iceberg tabl

Parquet tuning has always been like that, ever since it first came out in 2013. I worry with Iceberg that people think it's just a case of "use an Iceberg table in Snowflake" and boom, amazingly fast querying of data in S3!