ackratos · December 10, 2017 14:45 · ackratos · Jan 15, 2018
diff --git a/gistfile1.txt b/gistfile1.txt
 Could you please post your method for generating those benchmark results. The error margins appears too high to draw conclusions, so perhaps you ran with insufficient warmup, iterations etc? Did you build the library with the optimizer enabled? (sbt setupPublishCore dist/mkPack will do this.)

 Microbenchmarks should only call the code under test, so please remove the assertions (they are still run in test/junit/scala/util/matching/RegexTest.scala, the source of the examples you are benchmarking.

 What was the primary motivation to create a custom benchmark runner for this test? I tend to enable profilers interactively jmh:run -prof gc ... in the SBT shell.
	Could you please post your method for generating those benchmark results. The error margins appears too high to draw conclusions, so perhaps you ran with insufficient warmup, iterations etc? Did you build the library with the optimizer enabled? (sbt setupPublishCore dist/mkPack will do this.)

	Microbenchmarks should only call the code under test, so please remove the assertions (they are still run in test/junit/scala/util/matching/RegexTest.scala, the source of the examples you are benchmarking.

	What was the primary motivation to create a custom benchmark runner for this test? I tend to enable profilers interactively jmh:run -prof gc ... in the SBT shell.