Re: A new pseudo-disk for Apple 2.

[Linards Ticmanis <ticmanis@gmx.de>]

Alex Freed schrieb:

> In theory my floppy emulation can be near perfect in a sense that
> half-tracks and even
> quarter tracks can be emulated. But more involved firmware is needed and a
> new
> format for nibbles that indicates half-tracks, 36 an 40 uS sync bytes etc.

As a "nibble format" I'd suggest simply storing a 36 ï¿œS sync in 9 bits
and a 40 ï¿œS sync in 10 bits (i.e. just ignore byte boundaries and store
a raw bitstream), this is the most universal approach I think. Then just
duplicate the Woz Machine logic (not all that difficult if you have a
schematic of the old Disk II controller). Add a per-track bit count and
you should be fairly universal.

As for the disk selector: A NMI generating pin is an almost universal
approach - but not quite since a NMI writes three bytes to the stack,
which might contain useful data, though it's not all that likely that
they do. You'd really need to freeze the host CPU, DMA out the stack
page, and probably the zero page too, and then cause an NMI, forcing the
system to read *your* NMI vector instead of the one in ROM or RAM (can
you do that on the Apple?)

Problem with this: On a II/II+ you can't reliably freeze the CPU for
more than about 5 cycles, else it'll lose state.


-- 
Linards Ticmanis