PE> Would you like to restate your position on sectors?

RS> Sectors are normally numbered form a
RS> base of 1 for various historical reasons.

PE>> Do you think that sectors are:

PE>> A) 64 sectors, numbered 0 to 63

PE>> B) 64 sectors, numbered 1 to 64

PE>> C) 63 sectors, numbered 0 to 62

PE>> D) 63 sectors, numbered 1 to 63

PE>> Last thing I remembered you said A.

RS> Nope, B.

PE> Ok, close enough.  What evidence do you have of this?

The most obvious is sector editors where you specify the
sector in terms of head, cylinder and sector. The only one
that starts from 1 is the sector. LIke Nortons DiskEdit.

PE> I have been using D in PDOS, and it appears to be working.

Depends on how you decide it 'appears to be working',
if you dont rigorously test for the missing last sector....

Its also complicated by what level you are doing the direct access at.
There is lots of translation and faking with a modern IDE drive. You
dont actually have a fixed number of sectors per track anymore, that
varys in bands across the platter and that is faked up into something
completely different at the level of the commands down the cable to the
drive. Even more dramatically if the drive is being used in LBA mode.

There is usually reverse faking in the bios too so you can
make requests using cylinder head and sector numbers, thats
converted into an LBA block number for the command over the
cable to the drive, and thats then converted back into physical
cylinder head and sector numbers in the drive itself.

All gets rather complicated.
@EOT:

---