No, it uses PD.SYS... and the actual code which does the hardware control is pretty much in ROM... the PD driver calls the ROM, and the ROM knows about RT-11 queue elements. It returns with either success or failure. The actual memory layout is 60kb with a 4kb IOpage. It does not have a memory management unit, so RT11SB and FB are the only ones which run on it. Megan

No, the CPU board, memory and terminal interface board are a unit of 3 boards connected together. The disk interface is a separate board. I would not say no possibility, since I'm sure someone with the skill could hack something together, even if it was to take replace the floppy interface board with something that looked like a floppy to the CPU, but was actually a hard drive. But, if you are asking about any designed-in way of expanding with a hard drive -- no. The PDT is a bounded system * (* Bounded system - a system which would be usable if it could be expanded, but can't because it is bounded). There is no bus. The boards don't plug in anyplace, they are simply attached back-to-back (with the exception of the floppy interface board, which has a wire/ribbon cable (? - it has been awhile since I opened one up) connecting it. One Console port (standard CSR/Vector), three async serial ports (I forget which range they are -- I can check), one printer port, and one modem port which is either sync or async To accurately emulate it, someone would need to dump the ROM code and use it to process the RT requests. Another gotcha in the PDT is the slooowwwww IOpage accesses... If you do a word access, it takes 90ms (if I remember correctly), but a byte access takes 180ms. Also, there is something about read-modify-write instructions, you don't want to use lots of them on the PDT in interrupt code... I seem to remember code somewhere which has PDT conditionals to govern whether the code uses MOVs or BISs, the latter being avoided on the PDT. Megan