>Um:
>	(1) that is NOT "Nway autonegotiation" (as defined by IEEE 802.3u)

Yes, Jason. Well done. That's  what I *said*.


>	(2) that FSM is _precisely_ what our de driver does right now.
>	    It also happens to wedge the driver in OACTIVE from time
>	    to time.


Yes, we have an FSm. It also has bugs.

But what I heard first-hand from Matt Thomas was that DEC had a
different, smarter, better, and patented FSM.  
Whether the FSM our de driver now has is the same one Matt referred
to, i have no idea, nor do I know how I could tell.

BTW, we could also use a similar FSM for some other drivers, like the
3c595 and 3c597 -- if they dont have such an FSM yet.