Quote Originally Posted by Merri
Heh, should check other people's code and not trust it: I copy-pasted plenderj's code without too much checking on it I did wonder at one point why I had to have + 1 in my code. I guess I also miscounted how many items plenderj's code returned - or then plenderj's code has that weird flaw. Haven't checked. Luckily an easy fix. Edit plenderj, be your name cursed

Here are some results on my machine.

1 000 000 items:
- 25 ms yours (1.0)
- 45 ms mine (1.8)
- 165 ms plenderj (6.6)

10 000 000 items:
- 266 ms (1.00)
- 422 ms (1.58)
- 1688 ms (6.34)

100 000 000 items:
- 2656 ms (1.00)
- 4468 ms (1.68)
- 16906 ms (6.36)


In case interested: AMD Sempron 2800+, 1024 MB @ 333MHz
hrm... try looping it

Could be just a dif between intel an amd though