On Feb 5, 3:23=A0am, Andy McFadden  wrote:
> In comp.sys.apple2 mdj  wrote:
>
> > Unless such a format can allow for a bitream that records every
> > readable flux transition in a continuous track at every possible track
> > position, there's no way to guarantee it will not prove inadequate for
> > some software title or another :(
>
> The canonical example is a disk with physical damage; I think Mask of the
> Sun used this approach. =A0The disk has a tiny pinhole bored in it. =A0Yo=
u
> get inconsistent data back when reading it, and you will certainly never
> read back what you write.

The only viable approach in these cases is some kind of meta-data. I
like the idea of a NIB (that is nibble-based image) with inline meta
data for timing data or other anomalies. We certainly have the cycles
to spare in emulators.

Capturing the data is the real problem. I've experimented with some
custom sequencer PROMs, but it's not easy to get the Disk II
controller to do anything except exactly what Woz designed it for. ;-)

Cheers,
Nick.
--- SBBSecho 2.12-Win32