@dmiranda i've done some quick benchmaks with this tool: https://github.com/renaissance-benchmarks/renaissance using java 8 update 20 (i couldn't find update 17), update 152 and update 291. test were run with the following commands, picking the best result for each iteration: rx-scrabble -r 16, akka-uct -r 4, fj-kmeans -r 4
i couldn't find any significant performance difference. everything seems within the margin of error. maybe with other applications or tasks there is difference, but for general data processing and I/O seems the same. results are as follows (respectively: rx-scrabble, akka-uct, fj-kmeans):
update 291: 485.814 ms, 49395.213 ms, 7072.282 ms
update 152: 486.348 ms, 48090.700 ms. 7086.101 ms
update 21: 488.517 ms, 51775.709 ms, 6663.593 ms
i guess the advantage of using newer updates is security enhancements and running modern applications that do require newer updates to run at all.