What's new in Caliper 1.0 Beta 1
May 10, 2013
- Benchmarks extend 'Benchmark'.
- Can measure runtime and allocation.
- Runtime results from 0.5 version were subject to corruptions from factors like GC and JIT occurring mid-timing. 1.0 now detects and reports such events.
- The 0.5 version produced incomplete and possibly misleading console output. 1.0 does not.
- Caliper 1.0 uploads every run automatically, but an API key is still required to associate a run with a particular user.
- Can store the JSON result of every run in ~/.caliper/results/.
- The 0.5 version was configured using ~/.caliperrc while Caliper 1.0 uses ~/.caliper/config.properties.
- Many bug fixes