Not sure but it seems like without much refactoring we may be able to fix the loss of precision by setting the unused positions to -1/disabled

case DS_ALG_RELWEIGHT: /* 11 - relative weight based distribution */
           hash = idx->rwlist[idx->rwlast];                             // check/skip  here
           idx->rwlast = (idx->rwlast + 1) % 100;
          break;

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or unsubscribe.