[fastasucan]

That would be a strange choice for reading RSS with Duckdb. On a more serious note, it says so in the blog:

>I wanted to see DuckDB in action. Since DuckDB can already read remote text, CSV, JSON, Parquet, and even Excel, it seemed natural to try replacing requests with read_text

And

>In the end, this wasn’t about building a reader or replacing RSS clients. It was about bending DuckDB in a direction it wasn’t designed for and seeing how far it could go. I like DuckDB because it makes small data fun again: quick to set up, light to use, and surprisingly flexible when you throw odd formats at it.