On Wed, Jun 07, 2000 at 11:30:25PM -0700, wire wrote:
> Hello, I have several Ditto Max tape cartiridges that were previously
> used under windows. They are the 3.5 GB tapes and were compressed with
> whatever method Iomega used for the Ditto Max drives. I do not know if it
> was hardware or software compression or the type. Anyway, I would like to
Software.
> retrieve the data off of these disks before I erase them. I have tried to
> use tar, which naturally failed, and I also attempted to use dd.
>
> dd reports this:
>
> # dd if=/dev/ftape of=ft
> dd: /dev/ftape: Input/output error
> 0+0 records in
> 0+0 records out
You need bs=10k as an additional option to ftape; however, if you want
the lot, then you should use /dev/rawqft0 (IIRC) to retrieve
byte-for-byte.
> #
>
>
> I understand from the documentation that it is not possible to read or
> write to the dos format, however, is it possible to grab the data byte
> for byte, ignoring the format.
>
> If it is not possible to do this without writing additional software,
> could someone point out a few good resources of technical documents on
> the dos/windows filesystem and other tips and information I will need. I
> am a fairly adept programmer, but I have not dealt with filesystem or
> hardware issues before.
This is precisely what I tried to do, although I haven't worked on it
for a while.
The specs you need are available at www.qic.org, IIRC, in particular
the following will be useful:
qic113g
qic122b
qic123d
qic130c
qic154
qic3010h
qic3020h
qic80n
(these are the ones I've got copies of and I managed to understand the
tape format using them, though some irrelevant ones may also be in the
list)
There's nothing that states that the tapes follow the QIC standards,
but they appear to do it fairly well, except for a number of problems.
You could start off by using vtblc to examine the volume table.
Firstly, the volume table's segments don't appear to always match the
actual location of the data; there's a cumulative off-by-one error.
Second, I couldn't get decompression to work successfully; my code
works fine with the sample given in the relevant spec (one of the
above ones), but fails within a few Kb (the sample is tiny, so it
could be my code; but the off-by-one error makes me suspicious).
If you're interested, send me a private email and I'll send you what
I've got - you can continue it if you wish.
However, it may just be easier to find someone with a Windows
computer, take the hardware there and transfer it by using Windows to
restore to the hard disk, and boot off a floppy to write a tar. I do
have a Linux boot disk lying around somewhere which I constructed to
have support for the Ditto Max, which I could send you an image of.
Robie.
--
Robie Basak <[EMAIL PROTECTED]>
PGP signature
