| You are probably running into the same problem I had: the disk may need to
| be freshly formatted. NetBSD's RAM disk kernel can't do this at present.

I've come to the conclusion that NetBSD-current for the Sun3 series is
unstable, at least when it comes to building from scratch.  For now I'll
just be content to use a dumb "dd" backup onto the second disk as a backup.

I tried a bit more experimentation but struck out.  If I carefully left
the actual first-stage boot block stuff (via the original "dd") yet did
a "newfs" using NetBSD-current, it would punt with an exception - at least
the label was acceptable.  But something is broken (or I am screwing up)
when it comes to a "new" copy of "bootxx."  I'll wait a while and try
again.  I noticed that bootxx & ufsboot seem to be built differently now
(some dd'ing to strip headers and such) and that installboot has a new
hidden option -h related to whether things have headers or not.  I tried
booting after using installboot with and without -h, either an exception
or a page failure.

As for the QIC business - when I tried a viper on a uVAX using a clone
SCSI board I found it would not do I/O to the drive - I think it is because
the QIC has some very very specific block format requirements (i.e. it will
not accept settable block size, but requires a fixed block size).

-- Ken