[Coco] DSKINI glitch while using physical drive 1 and SDC DOS 1.5

[Darren A]

On Sun, Sep 3, 2017 at 11:35 AM, Rietveld Rietveld wrote:

> So far I have just waited until the head stop stepping. That way I know
> the formatting is complete. But motor stays on even after stepping is
> done.   A reset and DIR1 confirms a successful DISINI format
>
>

I don’t believe it’s a drive motor shut-off problem but rather the
verification pass of DSKINI slowing to a crawl, making it appear as though
the computer has crashed.

It seems that the code patches in SDC-DOS are now consuming a bit too much
time and causing the optimal track verification routine to slow to the
worst-case timing.  If setting the High-Speed poke before doing the DSKINI
fixes the problem then that would confirm my suspicion.

Technical details:

The DSKINI verification pass was designed to be able to read all 18 sectors
on a track in no more than 2 revolutions of the disk.  To accomplish this,
the DSKCON routine must be called repeatedly at a rate fast enough to see
each consecutive sector (in interleave order) without any slipping by,
otherwise another full rotation is required for each sector missed.

If too much time is elapsing between calls to DSKCON for each and every
sector then it will require upwards of 18 revolutions of the disk to verify
a single track. That’s about 3.6 seconds per track and more than 2 minutes
to complete the full verification pass.


- Darren