3

In Caliper ~0.5, benchmark runs were accompanied by some human readable console output for each benchmark run, e.g.:

stringCount            benchmark     ns linear runtime
          2               Concat   73.3 =
          2   StringBuilderSized  126.6 =
          2 StringBuilderDefault  182.9 =
          2               PlusOp  163.9 =

         10               Concat 1949.4 =========
         10   StringBuilderSized  620.7 ==
         10 StringBuilderDefault 1236.0 =====
         10               PlusOp 1288.9 ======

This output doesn't seem to be available in Caliper 1.0, but I'm wondering if there is any way to restore it?

The text output is very useful as both immediate feedback (especially when repeatedly tuning the benchmark or the code being timed), as well as textual output that can be easily and permanently embedded into various text sources. The web app is superior for most purposes (and includes additional critical information, such as the deviation of the results) - but there are times where the text output would be appreciated.

4

1 回答 1

2

有一个问题跟踪此请求：问题 237

于 2013-08-22T21:09:02.943 回答