The entire premise here is really about time-series workloads where most operations are based on a timestamp. In those apps partition dropping has been a standard and recommended retention strategy for years. That's precisely why extensions like pg_partman and TimescaleDB exist. Given that context, the title feels more clickbaity, and could easily mislead readers into thinking this applies broadly to OLTP systems when it doesn't;