22 Oct
2012
22 Oct
'12
4:36 p.m.
On Mon, 22 Oct 2012, Chris Tofu wrote:
C: So in the absence of optical means to disclose flux
transitions, will
anything else do the trick? In my experience once a disk is rendered
unreadable, putting it in another drive or whatever resulted in no
difference in results. Why I never obtained a catweasel. And generally
(overwhelmingly!) I don't even say "lessee what's on this little disk
here" and pull up a directory or whatever. Nuh uh, I go strait to trying
to image it.
It depends on WHY it is "unreadable".
I have repaired some "parity" errors, and SOME "Sector Not Found"
errors
(a few bits out of whack in the sector header), using TE.EXE on the
Central Point Option Board (a predecessor of the CatWeasel and DicFerret)
Yes, repairs at that level are slow and tedious.
SOME aspects of that could be automated. TE identifies where it sees
"inaoppropriate" or "missing" clock bits. It would not be hard (no
more
than a few man-years of work) to go further and identify which of those
are in positions as to be probable deliberate (address marks), and for the
others to try the various combinations of possible repairs to see which,
if any, produce a track image that "meets standards".
Therefore, once the flux-transition stream from the flux-transition board
(catweasel, discferret, option board, OR Chris' Optical Flux Transition
Finder) has been processed to produce a FM, GCR, MFM, MMFM, etc. bit stream,
that additional error identification and correction code could fix some of
the flaws.
THEN the code to separate sectors.
THEN file system code to extract files.
I don't think that we are ready to try to do it real-time to make a
"mountable" optical magnetic floppy drive (not to be confused with
Magneto-Optical!), BUT, a project of resurrecting a damaged floppy
disk MIGHT BE possible.
--
Grumpy Ol' Fred cisin at xenosoft.com