tisonkun opened a new pull request, #323: URL: https://github.com/apache/datasketches-rust/pull/323
Replace Divan's per-operation measurements with fixed workloads sized around hundreds of milliseconds, following [Benchmark In Milliseconds](https://matklad.github.io/2026/10/05/benchmark-milliseconds.html). Each revision runs the same amount of work, so improvements reduce elapsed time instead of changing the workload size. The runner measures fixture construction through cleanup and reports integer-millisecond medians, full ranges, and optional raw samples. XOR comparisons share round counts and rotate execution order across repetitions. Matched-build cases validate the resulting seeds directly, preserving equal construction attempt counts without building another implementation inside the measurement. All 279 existing operation, input, and implementation combinations are retained across 207 scenarios. Add filtering, exact selection, repetition and round overrides, listing, compile-only, and smoke-test support through `cargo x bench`. Document timing boundaries and baseline maintenance, and run smoke tests in CI without timing assertions. These measurements establish a new baseline because setup and destruction are now included. Validation: - `cargo x check` - `cargo x test` - `cargo x lint` in a clean source checkout - `cargo x bench -- --test`: all 279 combinations pass - `cargo +1.86.0 check --package benchmarks --all-targets --locked` - Three-repetition timing runs, including matched XOR comparisons and CPC serialization -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
