PE> packet had messages 7 and 8 corrupted (from an address 0/95).

PE> if you think you have found a PQWK bug, send me an
PE> archive including everything needed to demonstrate the bug.

Still not at all clear what caused that. I've just uploaded BAD_PAUL.ZIP
which contains a pair of REPs with just two messages in each, one which
produces a bad PKT on that address detail, the other which does not. The
only difference between the two REPs is the order of the pair of messages.
@EOT:

---