Re: Disk II Phase Question

["Jplcsch" <jplcsch@wanadoo.fr>]

"Michael J. Mahon" <mjmahon@aol.com> wrote in message
20050116110701.11015.00000032@mb-m18.aol.com">news:20050116110701.11015.00000032@mb-m18.aol.com...
> Jplcsch wrote:
>
> >Look the routine of  flight simulator II, it's really the maximum you can
> >put on a track (18 pages with a small amount for the translation table!)
> >on look the track with CIA or any other utility
> >to help you:
> >Syncro are $FE
> >start of track $D5 $AA $D5
> >after you found a transalation table
> >a close byte $94
> >the marker of every sector (just one) $D5
> >18 pages  (if you unassemble the read routine you'll found a loop with
$12
> >to test the nb of sectors read on the track)
> >checksum
> >close markers for the track $DE $AA
>
> I'm not sure what you mean by a "translation table".  The nibble
> translation table is usually generated, not written on the disk.
> If it were written, it would be 256 bytes long.
>
> Are you refering to something else?
>
> And is there no check byte?
>
> I note that the 18-page format you describe above requires 6172
> nibbles (of which perhaps 5-10 have two extra bits).  That works
> out to 193.5 milliseconds, or a little more than 3% less than 200 ms.,
> meaning the writing drive could be a little more than 3% fast and
> still work.
>
> Since the read routine can handle a writing drive at least 5% slow
> and a reading drive at least 5% fast, there is room for one more
> "page" of nibbles on a track if the writing drive is set 5% slow.
> (Another page would need only 5.5% more nibbles on a track.)
>
> Such a format could only be written a track at a time, but would
> provide a 19% disk capacity boost over the standard format.  ;-)
>
> -michael
>
> New Applesoft BASIC interface for NadaNet networking!
> Home page:  http://members.aol.com/MJMahon/


You're right but here it was a real transalation table but length was $40 (a
routine put in a correct order after. It was to do a mess when you try to
read a "sector" directly on the disk and try to unassemble with some
specific tools, why microsoft do that ? a crypt'll be better).

I know it's incredible but it's easy to verify. Just have a  look on the
read routine on flight simulator, it's easy to extract from the original
disk.

You're right for the speed of the drives.

The problem with this oversize if you'll use with a custom DOS is that you
must read one track to pick up a sector after in the loading  buffer (after
translation). Of course it's nice to put a program on this type of format
but not for a filesystem because to must read and write one track to modify
one sector only . not really funny...