On Wed, Jun 14, 2000 at 05:34:04PM -0800, Civileme wrote:

-> the defrag.  Note that Microsoft does not offer a defragmenter for NTFS...  It
-> is supposedly a more efficient system but who knows since they don't pass out
-> the specifications....

NTFS is more efficient than FAT32, but then it was designed to be used on
a hard drive, not a 180K floppy like FAT. I can't say how it compares to
ext2fs, though.

W2K comes with a defragmenter, I believe. It does provide the API calls
(via FSCTLs) to roll your own defragmenter, for both FAT32 and NTFS.


-> Windows is based on a close link between the physical storage on the disk and
-> the logical storage there.  ext2fs is at a higher level of abstraction,
-> deliberately fragmenting the free space on the disk so files are generally
-> unfragmented..  Recent improvements have made ext2fs even better.  rpmfind.net
-> reports success with the journaling ext3 fs, and you can see the Reiserfs is
-> now offered in linux-mandrake.  None of those filesystems really need
-> defragmenters.

Right. Defragmentation is not a necessity, it is a fix for a badly
designed file system, a kludge upon a kludge. So the fact of the API calls
for defrag and the defragmenter (if it is there) in W2K is an admission
that NTFS does not scale well from small hard drives (>100 MB, say), to
modern hard drive sizes.

One redeeming feature of the NTFS and FAT defragmenters in W2K is that
they are designed to be used while the file system is mounted and in
use. You can even move directories and files that the OS has open. That's
nice, but since it is possible to design file systems that don't require
defragmentation, why not do that instead?

-- 

                -- C^2

No windows were crashed in the making of this email.

Looking for fine software and/or web pages?
http://w3.trib.com/~ccurley

PGP signature

Reply via email to