>> Actually, you can use 36-bit SIMMs, as well.
> Yes, the hardware FAQ claimed this was possible -- just can't mix and
> match 33/36...

You can't?  I've never had trouble playing mix-&-match; it always just
uses the narrowest width of anything I've got in there.

Unless you mean that there's a problem with the <IPX,33,36> triple, and
I just haven't happened to try that particular one, which certainly
could be.

>> In fact, if you turn on the debugging software switch (something
>> like 'setenv debug-switch? true'

"setenv diag-switch? true", I think it is.

>> at the "ok" PROM prompt), when the machine resets, you will see RAM
>> tests done, and then a message saying something like "switching to
>> 33 bit parity mode" or "switching to 36 bit parity mode."
> Ah, cool!

Yes, it is.  I've used this mode as a memory sizing tool - when I have
a 72-pin stick of unknown size and width, I set diag-switch? true, pop
it in, and watch the messages.  If the stick is bad, it sometimes dies;
putting good 4Mx36 sticks on either side of it will usually make it
not-hang, reducing the failure mode to an error from the memory
selftest.

> Well, one can argue that if you *need* parity, you're already dealing
> with a flakey system/design!  :>

True.  But doing parity allows to at least find out that your system
has gone flaky, which is arguably better than silently getting bad
bits.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B