Remove pgrust: the entry is cheating (per-query kernels and result caching) - #2165
Summary
A GitHub pull request discusses removing a design that cheats benchmark results in ClickBench by using per-query kernels and a results cache. It explains the rationale, how the cache and tuned paths distort benchmarks, and notes that non-conforming queries are refused. The post highlights issues around benchmarking integrity in open-source projects and the impact on reproducibility.