Add Polars reference in performance section

Author: kingkongbot

src/ezmiller/relaunching_tablecloth_time.clj

@@ -45,6 +45,9 @@
 ;; tree-based index is pure overhead — you'd rebuild it constantly. As Chris
 ;; Nuernberger (author of tech.ml.dataset) put it: "Just sorting the dataset and
 ;; using binary search will outperform most/all tree structures in this scenario."
+;; (Notably, [Polars](https://pola.rs/) — the Rust-based DataFrame library gaining
+;; traction as a Pandas alternative — reached the same conclusion and has no index
+;; by design.)
 ;;
 ;; **On convenience:** The index adds implicit state threaded through your data.
 ;; Tablecloth's API avoids this — you always say which columns you're operating on.