Hello,
I'd like to introduce Native Histograms in the project, so histograms are are also exposed without all the limitations of traditional histograms (cardinality problems, bad bucket layout, etc.).
Adopting them provides a way to use them in an opt-in manner, and users can also scrape both old and new histograms.