Better ORC support plan?

[Xinyu Zeng]

Hi, I was wondering if there is a plan to improve the ORC reading
performance to the same level as Parquet? e.g. parallel reading across
columns, filter pushdown

Thanks,
Xinyu