From: David Brown on
Cronos wrote:
> Rod Speed wrote:
>> Doesnt happen with defragging, because the absolute vast bulk of linear
>> access to very large files is with media files where it takes EXACTLY
>> the same time to play the file whether its fragmented or not.
> Not to open or move the file though. That's were defrag comes into play
> and not FPS.

In theory, yes - in practice, no.

This is really a difficult concept for you to grasp, isn't it? It is
/correct/ that it is faster to read an unfragmented file than a
fragmented one. But the difference is so small that it is /irrelevant/
in almost all situations.

When copying a large file, it is correct that an unfragmented file will
copy faster. But that difference will be drowned in other effects, such
as where on the disk the file is located, or what the machine is also
doing at the time. Wiggling the mouse will probably cause more effect.

I did a fragmentation analysis on one of my disks - I have a file that
is in regular use (it's a virtual box virtual harddrive) that has
850,000 fragments. If fragmentation was relevant, that would really
crawl - in reality, the virtual box works perfectly well.

The report also reveals how astoundingly bad Windows is at generating
fragmented files. I have one file on the disk that is 8 MB in size, and
1863 fragments. This file has never been opened (certainly not for
writing) since it was originally copied onto the disk. This makes it
clear to me how pointless defragging really is - any conceivable gains
will be quickly lost by Windows' hopeless allocator.
From: David Brown on
Cronos wrote:
> David Brown wrote:
>> First off, everything is /never/ equal. Secondly, even if everything
>> else /were/ equal, who would notice or care? Unless you are in a car
>> race, a few percent longer or shorter on the journey is irrelevant.
>> I am not claiming that defragmenting has no effect - just that in the
>> great majority of cases, it has no /relevant/ or /noticeable/ effect.
> Then I take it you never upgrade your PC because it is always fast
> enough for you.

It is in fact almost correct that I never upgrade a PC - though I have
occasionally added more memory or an extra disk. It is usually far more
effective to buy a new machine when the old one is no longer sufficient
for modern software.

But when I buy a new machine - typically after five years or so - I buy
one that is /significantly/ faster and more powerful than the old one.
That makes a real difference.

Defragging in the hope of speeding up a machine is like the people who
pay most of the price of a good PC to upgrade their 3.4 GHz processor to
3.6 GHz.
From: David Brown on
Cronos wrote:
> David Brown wrote:
>> I suppose you are implying that I am parroting from Rod's posts? I
>> /do/ happen to agree with him in this matter - he is not wrong /all/
>> the time.
> Anyone who spends all day posting in this forum has a few loose screws
> so if that is who you choose as your hero then I guess you have a few
> loose screws upstairs also.

Did you read that, Rod? Apparently, you are my hero!
From: Cronos on
Rod Speed wrote:

> And only a fool upgrades from one to the other, fuckwit.

That is besides the point I was making but I guess the point was way
over your head.
From: Cronos on
Rod Speed wrote:

> No one does that.

Oh, yes, I forgot that you like to troll the home build hardware group
on the rare occasion too. The fact remains though that you spend way
more time than is healthy in this specific group.