On Mon, Jan 05, 2004 at 07:58:12AM -0800, Kenward Vaughan wrote: > On Mon, Jan 05, 2004 at 09:49:16AM +0100, Wilko Fokken wrote: > > On Sat, Jan 03, 2004 at 04:52:06PM -0500, Andy Firman wrote: > > > > > > Hello. I am not a hard drive expert and need some help in > > > understanding a few things. > > > > > > First, what is the difference between fdisk and cfdisk, > > > other cfdisk being curses based? > > > > > > > I read somewhere that 'cfdisk' should be better in defining partition > > borders fitting to those virtual cylinders and sectors of IDE drives. > > > > With my 8 years old computer, I found following the Win95 FDISK by > > cfdisk in order to partition the rest of my drives for Linux to be less > > error-prone than working with fdisk. > ... > > My own experience with M$ systems stems from keeping Wintendo 98 on the > drive for Freespace. Couldn't get that OS to not croak or attempt to > change the partition size--until I discovered it couldn't handle an > "oddly-sized" partition coming out of cfdisk (eg. 4224.5 Gb -- perhaps > anything not fitting a perfect marriage between block size and total > space?). > > Anyway, my thought is to make sure you pick a nice whole number of Gb > (a nice binary multiple or multiples added together). M$ may cry > otherwise. It's fussy about partitions ending on cylinder boundaries, and it seems to come up with its own ideas about what CHS geometry to simulate which don't necessarily correspond with what Linux would come up with given the drive in a blank state. If you define the M$ partitions first, with M$ FDISK, then cfdisk (at least) is smart enough to figure out what CHS geometry M$ FDISK has used and conform to that. -- Pigeon Be kind to pigeons Get my GPG key here: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x21C61F7F
Attachment:
pgpGGn7ideXHD.pgp
Description: PGP signature