Re: Questions on Disk ][ error checking

["Michael J. Mahon" <mjmahon@aol.com>]

Zardo Zapp wrote:
> dog_cow@macgui.com (D Finnigan) wrote:
>
>
> > I have a question about the error-checking mechanism used on the Disk ][.
> >
> > As I understand it, checksums are used to verify that the data read from the
> > disk are valid and error-free. However, would there be a way to skip these
> > checks? Is there any software that does so?
> >
> > I ask because I have more than one disk which has some data and programs
> > which can be read successfully, and some data which cannot. These disks are
> > not copy-protected. Would it be worth my time to make a disk image of a disk
> > with only partially good data? Is this even possible?
>
>
> I worked with Sheppy a while back to get a version of ImageMaker that
> can do this. It worked well for me.

ProSel's volume copy will retry reading a bad sector/block several
times, then ofer the choice to continue re-trying or ignore it (and
write a block of zeroes).

> Yes, it's worthwhile. The corruption may not be on an area of the disk
> you care about, and so you can salvage the rest.
>
> What would be nice, though, is to get a list of the damaged sectors so
> you know where the trouble may be. So far, I don't have a good way to
> do that.

You can use Cpoy II Plus' "disk verify" functiion to get a list of
"bad" sectors, and the "map files" function to find which files they
lie within, if any.

> > Also, would there ever be a circumstance where the disk has become damaged
> > in an area such that the Disk ][ cannot tell whether a bit is 1 or 0 and
> > thus refuses to read that disk?
>
>
> The hardware isn't that sophisticated. It'll read a 1 or a 0 no matter
> what - whether that's what supposed to be on the disk is anothe story.
> The checksum is only 8 bits, so if the disk is so bad that you get
> essentially random data, then on average 1 out of 256 reads the
> checksum will still match. In other words, the checksum only does a
> fair job of detecting data corruption. 

But in the more usual case of a few errors, it is quite effective.

-michael

NadaNet and AppleCrate II: parallel computing for Apple II computers!
Home page: http://home.comcast.net/~mjmahon

"The wastebasket is our most important design
tool--and it's seriously underused."