My first reaction is that the most likely problem is
the
RT-11 image on the emulated TU-58.
Thanks for the suggestions. As an update, I've done
a little more investigating. I ended up installing
tu58em and verifying that the tape image was good
and that I could boot using tu58em. Last night, I
did a dump of tu58em's debugging output and manually
issued the same set of reads to my emulator. I verified
that all the packets had the same checksums, the same
first few and last few data bytes, and that the end
packets all matched. So it's looking like it's coming
down to either something related to timing or there's
something I'm not doing right with respect to breaks
or inits. In looking over the RT-11 boot code, I'm
now pretty sure that things break down when it shifts
from using the boot version of read to the read
interface
the monitor uses. I"m thinking my next step is to interpose
a machine with 2 serial ports and record the whole protocol
when it fails and see if that shows anything.
BLS