>On this thread, I need to ask something.
>
>I have a DS5100 that I'm getting more and more miffed at, cuz DEC's
>policies for licensing Ultrix are neanderthal at best.

[[what would a NetBSD port require?]]

We'd need something called a "Functional Specification", like
the on-line PostSCript functional specifications for the 3100
and 5000/200 on gatekeeper.dec.com (I forget exactly where; if
they're gone, I have copies.)

The kind of info that's required is enough info to write drivers for
the 5100.  That means, what kind of baseboard devices are present
(ethernet chip, scsi? chip, Qbus interface(?), framebuffer,
serial/console ports, non-SCSI disk controller(??)) and at what
addresses they're mapped; how to force pending writes from the CPU out
to I/O devices; things like that.

Ted Lemon probably knows more about this than I do.  I've seen a 5400
and a 5500.  I have no idea what a 5100 is. Does it have a
turbochannel, or a Q-bus, or no bus at all?  If it's more like, say, a
5400 than a 3100, then the porting effort is likely to be
considerable. If there's any chips in common with CVAX-based machines,
then collaborating on device drivers with the NetBSD/vax port may be
worthwhile,

I'd also have to say that the DECstation machines here at Stanford --
which don't include a 5100 -- are becoming less of a priority for me,
and so NetBSD hacking is becoming less of a priority.