Re: EDD + CARD

["Charlie" <charlied@NOSPAMbboard.com>]

"Michael J. Mahon" <mjmahon@aol.com> wrote in message
20020727024633.06794.00001121@mb-fc.aol.com">news:20020727024633.06794.00001121@mb-fc.aol.com...
> Matthew Russotto replied:
>
> >In article <20020726152105.06824.00001292@mb-fc.aol.com>,
> >Michael J. Mahon <mjmahon@aol.com> wrote:
> >>
> >>Interesting.  The authors have apparently developed their own
> >>lexicon for the Disk ][ formats, quite different from the conventional
> >>one (perhaps invented for "Beneath Apple DOS"?).  For example,
> >>their "timing bytes" are usually referred to as "self-sync bytes" (or
> >>"sync bytes", or even "sync nibbles", since they are not data bytes).
> >
> >Beneath Apple DOS gets the lowest-level totally wrong, I think --
> >they talk of clock bits between each data bit, which if I understand
> >correctly simply aren't there; that would be an FM scheme.
> >
> >Not that this has anything to do with self-sync bytes, which they do get
> >right.
>
> Yes, earlier today I pulled out "Beneath Apple DOS" to check what
> they called things, and was amazed to find that they described the
> low-level as FM encoding, with clock 1's between each data bit!
> This had the surprising effect of doubling the read pulse rate to
> 500KHz--which it seems would have tipped them off to the error.
>
> They must also have noticed a certain illogic in the GCR pattern
> constraints (required for accurate bit timing recovery), when FM
> would completely remove any constraints.  (In fact, the "4+4"
> coding used for addresses is exactly FM, itself.)
>
> I had an old note in there correcting my copy, and I checked that
> by "Beneath Apple ProDOS" they had it right, though their read
> waveform drawings are still a little peculiar, with little circles on the
> baseline to mark the "edges" of the bit cells--right where their clock
> pulses used to be.  ;-)

Later editions of "Beneath Apple DOS" also corrected these errors.  My copy
(fifth printing March,1983) shows little dots between the bit cells.  They
define a bit cell as the distance the diskette moves in 4 machine cycles or
about 4 microseconds.
I had an earlier edition but a friend "borrowed" it for a "few days" back in
1982. I'll check it for errors when he returns it.

Charlie