Hello Dave!

  DM> Umm, where did these specs come from again?

A whole bunch of different places, which can only add to my suspicion
that there are errors in there.

  DM> If memory serves (and I'm pretty sure it does in this case) the RD32 was a
    > Seagate ST251, not an ST251-1.  Not that there's THAT much difference
    > (40ms vs. 28ms)..

I don't know which the RD32 was, but I've a recollection that the 'physical
geometry' was different between the ST251 and ST251-1 as well as the speed.
Sadly I don't have the relevent notes any more.

  DM> But yes, I suppose the use of an MFM drive to load the kernel from would
    > do fine, though I've never tried this myself on a '2000.

Patching the firmware would certainly be a more elegant solution, but there's a
little bit of me just dying to know whether I could make my approach work.   :-)

  DM> I'm doing that with an AlphaPC 164...I'm loading my kernel from a crap IDE
    > drive, which then mounts its filesystems from a real disk (a big zippy IBM
    > 36GB LVD drive on a big zippy Qlogic QLA-1080). The scheme works fine with
    > that hardware.

It's possible to get flash modules that look electrically like ATA drives,
depending on the application I might use one of those to bring up an AlphaPC
164.  I doubt I'd find an MFM equivalent for the MicroVAX 2000 though :-/

Regards,
  - Andy Ball.