User talk:Thundergnat: Difference between revisions
→Anti-primes - concurrency failure?: Probably worth submitting as a bug
SqrtNegInf (talk | contribs) |
Thundergnat (talk | contribs) (→Anti-primes - concurrency failure?: Probably worth submitting as a bug) |
||
Line 391:
:::Using those revisions, and here's my experience on Linux: Under normal load 'Anti-primes' usually works when run as a foreground process, sometimes fails when run in the background. Under heavy system load, sometimes fails in the foreground, and almost always fails when run in the background. Specific details of my test setup, procedures and results [https://gist.github.com/SqrtNegInf/a7838ffa90a3cbb3031dae15828bd460 in this gist.]
::::It's probably worth submitting this as a bug, though I suspect it is going to be hard to trace. If it was using .race, well, with .race, you rolls your dice and you takes your chances... but .hyper is specced to always return items in the correct order; if it isn't, that's unspecced behaviour. --[[User:Thundergnat|Thundergnat]] ([[User talk:Thundergnat|talk]]) 14:32, 13 January 2019 (UTC)
| |||