sample results

[boleslaw]

My results were similar from running your test 4770K/xubuntu/1.8.0_11-b12, posting here in case it helps. I like the java 8 api's but I always benchmark them and have to comment about why I did or didn't use them in each case. Thanks for posting this project, I didn't know about jmh.

Benchmark                                                           Mode  Samples     Score     Error  Units
c.e.n.b.ArticleOpsBenchmark.getAllJavaArticles_loop_benchmark       avgt       10  2454.833 ±  51.900  us/op
c.e.n.b.ArticleOpsBenchmark.getAllJavaArticles_stream_benchmark     avgt       10  2378.955 ± 230.910  us/op
c.e.n.b.ArticleOpsBenchmark.getDistinctTags_loop_benchmark          avgt       10  5123.317 ± 292.443  us/op
c.e.n.b.ArticleOpsBenchmark.getDistinctTags_stream_benchmark        avgt       10  6221.680 ± 346.158  us/op
c.e.n.b.ArticleOpsBenchmark.getFirstJavaArticle_loop_benchmark      avgt       10  2541.868 ± 172.443  us/op
c.e.n.b.ArticleOpsBenchmark.getFirstJavaArticle_stream_benchmark    avgt       10  3048.842 ± 114.207  us/op
c.e.n.b.ArticleOpsBenchmark.groupByAuthor_loop_benchmark            avgt       10  4372.353 ± 210.431  us/op
c.e.n.b.ArticleOpsBenchmark.groupByAuthor_stream_benchmark          avgt       10  4522.950 ± 279.239  us/op

[skinny85]

Hey, pretty cool. I can see that your results are even closer than mine (streams version of getDistinctTags is only 17% slower than the loop one - it was 24% in mine). Thanks for the feedback.