🚀 go-pugleaf

(Followups to "alt.dev.null"; if you must follow up, please consider
picking an even smaller set of groups than the set to which I've limited
my followup.)

>>I do know that the
>>Sun systems I've used since that VAX 750 have had miserable hardware
>>documentation; the situation is improving slightly only because Sun is
>>now using more third-party hardware in their systems, and you can buy the
>
>You can get documentation about Mbus & Sbus

But that's not enough to write your own OS for, say, a Sun-4m machine. 
There's plenty of hardware that the OS has to tweak that's not specified
by the Mbus or SBus (yes, that really *is* how they're capitalized, at
least according to the specs I have; no, I don't know why the "b" in
"bus" is capitalized differently) specifications.

>How about HP?  I'm not too familiar with their systems but I have heard their
>workstations use an EISA bus...

Well, the folks at the University of Utah *have* done a BSD port to
various HP's; however, I think that, at least at one point, the port
required that they lift some HP-UX code, and you needed a source license
to get the source to their port.

I.e., to fit Geoff's definition of "open", it's *not* enough just to
have an "open" peripheral bus.