The article is harsh for no reason. Pandas served us, for a lot of time and was the de-facto library for Data Scientists. It evolves. Polars was a rewrite following bets practices learnt while working with Pandas.

I do not understand the aggressiveness towards previous evolutionary steps. Who knows, may they come up with newer methods. I still use it. I also use DuckDB and Polars. DuckDB is more useful to me these days from either. Still Pandas has a place in my arsenal. Mostly because I know it better and I like it.

There is also libraries for doing type checking.