Paul Koning wrote:
>>>> "Brad" == Brad Parker <brad at heeltoe.com> writes:
Brad> Alexander Voropay wrote:
Brad> I love putr, but it requires dos.
Brad> I don't think the problem in interleave, as the resulting
Brad> images work perfectly with simh (as I said before). Simh
Brad> treats them as a big byte stream and does not deinterleave.
Brad> (plus, if you use the controller to read sectors by number, you
Brad> automagically deinterleave)
I'm not so sure about that. The RX01 driver I looked at did the
interleave and skew I described in software.
If you're running RT11 on SIMH to read your RX01 images, the RT11
driver would be doing any necessary block number translation. But if
you're trying to read the image directly then you have to do the
mapping.
paul
I believe Paul is exactly correct. I found this to be true also for XXDP
RX01/2 images (I am cloning a version of 'xxdpdir/diagdir' that also
can read/write RX01/2 media). The DEC floppy drivers (XXDP for sure,
probably RT11 also) interleaves/skews logical sectors to physical sectors.
SIMH uses physical track/sector addresses to access the image file, BUT
the mapping from logical to physical sector uses the interleaving function.
For all (?) other disk formats the logical to physical mapping is a simple
sequential function. But not for floppy images. You also have to know
whether
the underlying disk image is RX01 (128B sector) vs RX02 (256B sector).
The 'even/odd-skew-by-6' algorithm mentioned in a previous email is what
I found works for XXDP. It is probably the DEC 'standard' for RX01/02
drivers.
Don