Attention is currently required from: laforge, fixeria, pespin.

View Change

1 comment:

• Commit Message:

  • Patch Set #3, Line 12: Use OTC_SELECT

    Let me add that I don't know how it happened, but i overlooked some responses when writing the above.

    I appreciate the benchmark:

    the static __thread buffer implementation 1294870 times per second on my laptop, while the OTC_SELECT variant runs "only" 808880 times per second.

    It would also be interesting to compare whether there is a difference between the two variants when the logging is turned off: is there an impact by unused static buffers?

    I still don't get in general, why we should use a more expensive OTC_SELECT approach over a static buffer approach *without a clear requirement or reason*. It's not that it makes the code any easier to read or write.

    But clearly, it does, and clearly there are reasons for it, and requirements that OTC_SELECT meets better than static buffers do.
    Using OTC_SELECT helps to completely rule out several entire families of problems in code maintenance, <cue the previous long comment>

To view, visit change 33169. To unsubscribe, or for help writing mail filters, visit settings.