You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
By the looks of it, neither Scala 2.11 nor Scala 2.12 clearly dominate one another on re2s. Scala 2.11 is faster in cold mode, whereas Scala 2.12 is faster in hot mode. Let's include both in our benchmarks.
The text was updated successfully, but these errors were encountered:
This commit splits project `bench` into multiple projects:
`benchJavac18`, `benchRsc`, `benchScalac211` and `benchScalac212`.
This provides additional flexibility and fine-grained control over
target compiler versions.
The main motivation for this change is the fact that neither Scala 2.11
nor Scala 2.12 clearly dominate one another on re2s. Scala 2.11 is
faster in cold mode, whereas Scala 2.12 is faster in hot mode.
While refactoring the benchmarking infrastructure to accommodate this
change, I've also retired CliBench. It is confusing to have multiple
benchmarks that test the same thing (ColdXxxYyy and CliXxxYyy),
so I picked the one that's simpler, i.e. ColdXxxYyy powered by JMH.
RscNativeTypecheck is still powered by CliBench-like logic, but it has
been merged directly into that particular benchmark.
I am not completely happy with the astonishing degree of duplication
between benchScalac211 and benchScalac212, but my sbt skills don't
allow me to remove this duplication in a satisfying manner. (I'm aware
of ++, but I want Scala versions to be part of the project name, not
an optional prefix to an sbt invocation.)
This commit splits project `bench` into multiple projects:
`benchJavac18`, `benchRsc`, `benchScalac211` and `benchScalac212`.
This provides additional flexibility and fine-grained control over
target compiler versions.
The main motivation for this change is the fact that neither Scala 2.11
nor Scala 2.12 clearly dominate one another on re2s. Scala 2.11 is
faster in cold mode, whereas Scala 2.12 is faster in hot mode.
While refactoring the benchmarking infrastructure to accommodate this
change, I've also retired CliBench. It is confusing to have multiple
benchmarks that test the same thing (ColdXxxYyy and CliXxxYyy),
so I picked the one that's simpler, i.e. ColdXxxYyy powered by JMH.
RscNativeTypecheck is still powered by CliBench-like logic, but it has
been merged directly into that particular benchmark.
I am not completely happy with the astonishing degree of duplication
between benchScalac211 and benchScalac212, but my sbt skills don't
allow me to remove this duplication in a satisfying manner. (I'm aware
of ++, but I want Scala versions to be part of the project name, not
an optional prefix to an sbt invocation.)
By the looks of it, neither Scala 2.11 nor Scala 2.12 clearly dominate one another on re2s. Scala 2.11 is faster in cold mode, whereas Scala 2.12 is faster in hot mode. Let's include both in our benchmarks.
The text was updated successfully, but these errors were encountered: