diff --git a/tool/grind/benchmark.dart b/tool/grind/benchmark.dart index 00990aef6..781b1b704 100644 --- a/tool/grind/benchmark.dart +++ b/tool/grind/benchmark.dart @@ -101,16 +101,11 @@ benchmark() async { environment: {"SASS_LIBSASS_PATH": p.absolute(libsass)})); log(""); - var ruby = await cloneOrPull('https://github.com/sass/ruby-sass'); - log(""); - var libsassRevision = await _revision(libsass); var sasscRevision = await _revision(sassc); var dartSassRevision = await _revision('.'); - var rubySassRevision = await _revision(ruby); var gPlusPlusVersion = await _version("g++"); var nodeVersion = await _version("node"); - var rubyVersion = await _version("ruby"); var perf = File("perf.md").readAsStringSync(); perf = perf.replaceFirst(RegExp(r"This was tested against:\n\n[^]*?\n\n"), """ @@ -118,7 +113,6 @@ This was tested against: * libsass $libsassRevision and sassc $sasscRevision compiled with $gPlusPlusVersion. * Dart Sass $dartSassRevision on Dart $dartVersion and Node $nodeVersion. -* Ruby Sass $rubySassRevision on $rubyVersion. """); @@ -197,10 +191,6 @@ I ran five instances of each configuration and recorded the fastest time. await _benchmark("node", [p.join('build', 'npm', 'sass.js'), path]); buffer.writeln("* Dart Sass on Node.js: ${_formatTime(nodeTime)}"); - var rubyTime = await _benchmark( - "ruby", ["-I", p.join(ruby, 'lib'), p.join(ruby, 'bin', 'sass'), path]); - buffer.writeln("* Ruby Sass with a hot cache: ${_formatTime(rubyTime)}"); - buffer.writeln(); buffer.writeln('Based on these numbers, Dart Sass from a native executable ' 'is approximately:'); @@ -208,7 +198,6 @@ I ran five instances of each configuration and recorded the fastest time. buffer.writeln('* ${_compare(nativeExecutableTime, sasscTime)} libsass'); buffer.writeln( '* ${_compare(nativeExecutableTime, nodeTime)} Dart Sass on Node'); - buffer.writeln('* ${_compare(nativeExecutableTime, rubyTime)} Ruby Sass'); buffer.writeln(); log(''); } @@ -236,8 +225,8 @@ Future _version(String executable) async => Future _benchmark(String executable, List arguments) async { log("$executable ${arguments.join(' ')}"); - // Run the benchmark once without recording output to give Ruby Sass a hot - // cache and give other implementations a chance to warm up at the OS level. + // Run the benchmark once without recording output to give implementations a + // chance to warm up at the OS level. await _benchmarkOnce(executable, arguments); Duration lowest;