At 10:36 AM 6/4/2002 -0700, Bill Studenmund wrote:
>On Tue, 4 Jun 2002, Ignatios Souvatzis wrote:
>
>> I'm confused - aren't ffs (-like) block pointers in terms of the fragment
>> size?
>
>Yes they are.
>
>The questions are:
>
>1) are all internal settings in terms of fs blocks (fragments), or are
>some still in terms of the underlying sector size?
>
>2) what about other file systems that have even stricter requirements
>(like FAT file systems depend on the C/H/S settings of the underlying
>device; change them and the fs strictly-speaking needs to be
>re-formatted).

They do??  There's another monkey wrench... I thought C/H/S was only used
to convert a linear sector number to its 3-D coordinate for older drives.
What part of FAT uses C/H/S?  (Other than the partition table, which isn't
part of FAT anyway.)

If it will help, I think I have a spare 640MB MO disk I can use to format
with M$-Windows and see how it gets laid out.  (I also have a DVD-RAM
formatted under Windows, but that used a special driver with the UDF file
system.)

-----------------------
Trevin Beattie          "Do not meddle in the affairs of wizards,
trevin@xmission.com     for you are crunchy and good with ketchup."
      {:->                                     --unknown