You are viewing a plain text version of this content. The canonical link for it is here.
Posted to devnull@infra.apache.org by "leventov (GitHub)" <gi...@apache.org> on 2018/08/24 01:20:13 UTC
[GitHub] [incubator-druid] leventov opened pull request #6226:
Prohibit some Random usage patterns
- `Math.abs(random.nextInt())` doesn't guarantee positivity
- `random.nextInt() % x`
- `new Random()` should never be used, `ThreadLocalRandom.current()` or `SplittableRandom` instead.
I also replaced most `new Random(SOME_CONSTANT_SEED)` in benchmarks with `ThreadLocalRandom.current()` because benchmarks shouldn't generally be always identical, randomness between the runs could make results statistically more reliable.
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org
[GitHub] [incubator-druid] jon-wei closed pull request #6226:
Prohibit some Random usage patterns
Posted by "jon-wei (GitHub)" <gi...@apache.org>.
[ pull request closed by jon-wei ]
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org
[GitHub] [incubator-druid] leventov commented on issue #6226:
Prohibit some Random usage patterns
Posted by "leventov (GitHub)" <gi...@apache.org>.
@jon-wei fixed
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org
[GitHub] [incubator-druid] jon-wei commented on issue #6226: Prohibit
some Random usage patterns
Posted by "jon-wei (GitHub)" <gi...@apache.org>.
@leventov Can you fix conflicts?
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org
[GitHub] [incubator-druid] glasser commented on pull request #6226:
Prohibit some Random usage patterns
Posted by "glasser (GitHub)" <gi...@apache.org>.
Shouldn't this be Integer.SIZE rather than Integer.BYTES? Just doing unit analysis it seems wrong: (bytes / (bits/symbol)) = symbol * (bytes/bits) = symbol/8. This is giving single-character IDs for me, which seems wrong.
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org
[GitHub] [incubator-druid] jon-wei commented on pull request #6226:
Prohibit some Random usage patterns
Posted by "jon-wei (GitHub)" <gi...@apache.org>.
@glasser This was fixed in #6383
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org
[GitHub] [incubator-druid] glasser commented on pull request #6226:
Prohibit some Random usage patterns
Posted by "glasser (GitHub)" <gi...@apache.org>.
I guess that'll teach me to run off of a random commit on master :)
[ Full content available at: https://github.com/apache/incubator-druid/pull/6226 ]
This message was relayed via gitbox.apache.org for devnull@infra.apache.org