My feedback is that I don't have a lot, but I have some questions that
perhaps those of you more intimate with the various Vax architectures could
explore.

I happen to like the idea, I agree with Brian that perhaps vsbus and mvbus
don't capture the systems one might expect them to exist on. 

By the same token I can see Ragge's point that "mvbus" is infact simply the
place where all the possible busses for a smaller VAX might connect. (Q,
Uni, and internal)

Does anyone have a picture of how it might look (my VAX architecture book
thinks the 750 is a "new" vax :-)

And what about CPUs? At some point it would be useful to build up and ASMP
or even an SMP system and where would the CPUs connect?

So a given VAX would have 1 or more CPUs connected to the "xxbus" (those
are the cluster busses), and then that same vax might also have bus
adapters (Q, Uni-, etc) connected also to that cluster bus. Now the 6000
seems to do in-cabinet clustering whereas the uVAX does network based (NI?)
clustering? 

I think the bottom line is that if we can share more code its good. This
gets more platforms supported more quickly.

--Chuck