24 Mar
2005
24 Mar
'05
6:38 a.m.
On Wed, 23 Mar 2005, Eric Smith wrote:
Suppose you format a disk on a Disk II that is 10%
fast. Now you stick
the disk in a drive that is 10% slow, and write a sector. The 342
nibbles of the data field will take up the space on the disk that was
formerly used by 417 nibbles on the disk, or the 342 nibbles of the
data field plus 75 nibbles of the gap.
And I've ignored the header and trailer bytes, and the new self-sync
written in the gap and after the end, so it's actually worse than
that.
As it is, 80 nibbles of gap isn't quite enough to allow for +/-10% speed
variation and leave still leave enough self-sync bytes, but if they
reduced it, writing a sector might overwrite the address field of the
next sector, rendering that sector unreadable (even though the data
field of that sector would still be intact).
They couldn't fit 17 sectors without seriously compromising the
allowable speed tolerance.
If you're drive is spinning at 330RPM (10% fast) or 270RPM (10% slow) then
your disk drives are so out of whack that you should *EXPECT* to have
major problems.
Some copy-protected games crammed in more by writing
one giant sector
per track.
Right, but it shows it was possible. And there are tricks that one can
employ, such as on the fly testing of the drive speed to adjust DOS timing
loops, to ensure reliable writes at a more granular level. At least
that's the theory I'm going to set out to prove.
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]