You have not quoted me correctly - what I said was that the Ditto Max
cannot be run at its full rated speed without the overrun errors, but if
you restrict the speed you can have error-free operation. My own Ditto Max
drive runs on a Dash accelerator, restricted to 2,000 Mbit/s, without
errors or time wasting repositions. I suspect that I could run it faster,
but at least now it is stable and give me no grief.
I believe that the authors of FTape did evaluate the option of hacking
their code to better support the Ditto's unusually processor intensive
encoding and error checking protocol, but concluded that the main
trade-off was to sacrifice multi-tasking - something that they were not
prepared to do. If I recall, the lower speed problem is caused by the
specification of the floppy interface devices that the majority of
smaller/cheaper drives use. The fact that the Ditto hijacks this method of
working, and then accelerates the drive with a proprietary interface, is
from my understanding why the anomaly arises.
I set up a test-bed system (Pentium 75 with 32MB and 8GB of disk) to
experiment with FTape and the Ditto and found that this lowly platform
could support 2,000 MBit/s without errors, but no faster. So your system
ought to at least do the same, unless you're running on a 486.
It is quite possible to operate the Ditto Max with either the Dash
accelerator card OR directly onto the floppy bus - my test system was
configured for both modes of operation and worked well in either. The
setup with the Dash card was nearly double the speed of the floppy-direct
configuration.
With regard to patching, the FTape authors must answer your question.
Either that, or you must provide the patch yourself. Personally, I would
not be too interested in writing such a patch (even were I capable) when
the drives are known to be obsolete and will not be on the market for much
longer. Better to concentrate development effort on the newer hardware
which users are actually buying - like the OnStream device, or USB
connected devices.
> Original Message
Andy Corteen and Martin Jacobs answered to my question regarding the
"overrun error" of ditto max stating that these errors cannot be prevented
because of various items.
Now, don't you think that a real alternative to junk the ditto max could
be
to patch the ftape code in order to let it deals also with speeds lower
than
2Mb/s? If we were able to drive the to the streamer at 1Mb/s, I think that
the global speed factor would be enhanced, because of the reduced
number of retry due to overrun.
In fact, 1Mbit/s means more than 7.5 Mbytes/minute that is almost three
times the rate I measured using a theoretical 2Mbit/s setting.
Actually, if I set the parameter ft_fdc_rate_limit to 1000, the drive
won't
be recognized nor run. Do someone knows how to modify the software in
order
to let this value to be accepted?
<snip>
