12 Jan
1999
12 Jan
'99
4:55 p.m.
Fred Cisin wrote:
When going to the directory on a system without a
track 0 stop or track
0 switch, it DOES substantially reduce the average access.
But, if there IS a track 0 stop, or track 0 switch, then it is normally
possible to get to track 0 at full maximum velocity, rather than stepping
one track at a time, pausing (and maybe reading) before taking the next step.
That usually makes even the maximum seek to track 0 take even less than
half the time that track by track stepping would.
Sam Ismail wrote:
I don't know if this makes complete sense on the
Apple disk drives. When
you wanted to step, say, 5 tracks over, you massaged the stepper motor the
requisite number of times to get it to where you assumed it would be 5
tracks over. Then the DOS went about its business reading for the sector
it was looking for. If it didn't find it then it assumed a problem and
recalibrated (to track 0), then stepped back the requisite number of times
to get to the track you originally were headed for. This scheme resulted
in 99.99999% success in jumping from track to track. And the CATALOG
track was 17 (on a 34 track formatted disk).
Also, Apple didn't use the fixed step rates that were standard in the
industry. They accelerated and decelerated the carriage in order to
get faster seeks.
However, when they wanted to recalibrate the drive, instead of doing a
normal seek to track 0, they did an outward seek of 48 tracks. Since it
was only a 35 track drive, the positioner would hit the end stop and make
an annoying loud grinding noise.
So even though the Disk ][ had a track 0 stop, the use of track 17 for the
catalog was quite appropriate. The average time to seek to track 17 was
noticably lower than the average time to seek to track 0 (or the recalibrate
time).
Eric