🚀 go-pugleaf

In article koning@koning.enet.dec.com (Paul Koning) writes:
>Simple... PCM (the initialization protocol SMT runs between two ports) is
>a two-party protocol.  It would have to be DRASTICALLY different to be
>an N-party protocol, which is what you need to initialize the sort of
>topology you mentioned.  Actually, it's not clear that such a protocol
>could even be designed.
>
>People have experimented with these topologies, typically by not running
>any startup protocol at all but simply putting everything into active
>state right away.  With some luck that may work often enough to be 
>useful in experiments...

Hi Paul.  What's up?  In FDDI-2/SMT-2, people are talking a simplex 
protocol designed to connect SAS together like a bus.  The protocol calls
for a bypass (most likely electrical, to be used with TP-FDDI, but can 
be optical) for each SAS.  And the simplex PCM is a ONE way, three signal 
protocol (geez, not much of a protocol to it).  

The following is one possible configurations:

	T<=>BP<======>BP<=======>BP<=======>BP<=>T
	    |$        |$         |$         |$
	    |$        |$         |$         |$
	    SAS       SAS                   SAS

BP=Bypass, T=Termination Loop, |$=duplex cable with BP-on/off.

There are other possible configuration which connects to a concentrator
which supports simplex operation also.  The Oct 1992 X3T9.5 mailing
should show more detail info.  I think that the Simplex operation will be 
added to SMT-1 also.  (But then again, I can speak for the committee.)

---
| Michael Yip			Email: my@berlioz.nsc.com
| National Semiconductor Corp	Voice: (408) 721-5148
| 
| Of course, this is my own opinion!