Hi,
So, with a random chunk order, a factor of 2 was expected, which your figures confirm. The difference between the two patch variants should be much smaller.
Well i think with the former I did a little more downloading. So I can't make any conclusions from that. Then I have to do a more controlled test. But i have no idea how to do that with all the buffered IO, etc. -- Niek -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
