i think "SELECT * FROM my_df" is a convenience from the python module and how tightly integrated it is, but i can’t get this to replicate using the cli or dbeaver or datagrip.

and yes, being able to layer analytical sql on top of your csv/json/parquet/gpx/arrow (but not xml?) is the massive appeal of duckdb for a variety of reasons. it’s a paradigm shift for me as an old timer but it’s also suited my needs quite well over the past few years