I didn't attack anyone. My comment was directed at the thread in as a
whole and not Hussam. And right now I'm replying to THEfog because he
is the last reply but this post is directed at MAD_BEAST's comments
and to a lesser extent the peo who share his opinion. Now this is
going to be the last I comment on this subject and I'm going to try
and and phrase it more eloquently then my first attempt.

Now I'm not sure what you mean by axis, but I'm going to hazard guess
you mean the actuator arm (which I think the pivot point can be called
an axis). Assuming this is right my statement still hold true. The
drive heads write data from the outside of the platter toward the
centre. This explains why my 32g partition is many times faster then
my 400g partition.

Which brings me to your point. For arguments sake I listen to you and
move my swap file to the 2nd partition (situation 1). Right now my 2nd
partition is 90% full and well defragmented. Moving the the swapfile
now would most likely put it at the very end of the drive closest to
the spindle. Since my games are on the OS partition this would
maximize the amount of travel the head would have to do, thus
maximizing head latency and probably also rotational latencies.

Now assuming I had moved the swapfile while the partition was empty
(situation 2) this would have still put me past the end of the OS
partition (which is still not full as of this writing) so while I am
better off then situation 1 it will never be the same as sharing the
OS partition.

All of this assumes you have one hard drive (ie: most laptops, which
also would have these crappy graphics cards necessitating the super
human effort of the bravo team). If you have more then one hard drive
go ahead a move it. If you have IDE drives make sure your not moving
it to another drive on the same cable because you will still hose your
performance.

AS I said the process of using a small OS partition as the front of
the drive is called "short stroking". Now there are different ways to
do it, if you have consumer hardware just making a small first
partition is easy. Those of you who use hitach enterprise hardware
also have the niagara utility. In practice they both accomplish the
same thing but have key differences. Just making a small partition
allows you to use the rest of the drive, while a vendor utility
actually limit the usable portion of the drive. If you use a setup
like mine (2 partitions) you have to keep in mind that anytime you
access the second partition you will incur a performance penalty. But
since I use mine as an archive this is a penalty worth paying for the
use of the 400g+ space.

On Feb 7, 2:50 am, "THEfog ." <[email protected]> wrote:
> Nah its all right DanielPK, don't think of it again, I should have changed
> my post to be less hostile so yeah don't worry about it :)
>
> THEfog
>
> On 07/02/2010 5:19 PM, "DanielPK" <[email protected]> wrote:
>
> sorry if i miss-understood the subject...THEfog my bad man :) its just
> that you had and opinion and i had mine, and i should've respected
> it.....MAD_BEAST....you irresponsible bastard, without you we're
> nothing :))
>
> On Feb 7, 12:22 pm, MAD_BEAST <[email protected]> wrote:
>
> > Dont worry guys! life is too short f...
>
>

-- 
INTEL 9xx SOLDIERS SANS FRONTIERS

Reply via email to