Con Call Reminder - Protection for Asynchronous phases

Jim McGrath Jim.McGrath at
Mon Mar 1 10:04:39 PST 1999

* From the T10 Reflector (t10 at, posted by:
* Jim McGrath <Jim.McGrath at>

A reminder that there is a con call to review 99-199r0 (which should be
available on the T10 site), covering protection for information transferred
during the asynchronous information phases (COMMAND, MESSAGE, STATUS).
Logistically, you should call up 408-894-4600.  You need the meeting ID
number: 4504.  I would hope that we can finish up in 1 hour, although I have
scheduled the call itself for 2 hours.

The current revision of the document is technically complete, but does not
have the SPI-3 modifications.  I will modify the document to include those
modifications after the conference call.  The document (R1) will then be
ready for review next week.

On the conference call, I would like to start with a brief introduction on
the background for this proposal, and then review the technical contents.
The primary purpose of the con call is to allow for specific technical
feedback, particularly feedback that would allow for easier implementations.
At first it looked like a hard feature to implement, but John Lohmeyer's
contribution at the last working group meeting (using the upper 8 data bits)
is the key to allowing for an extremely straightforward implementation.
Quantum's contribution has been on working out the coding scheme and
specifics of the protocol (with feedback from a number of you).

I have received some private feedback questioning the need for this feature
at all.  This feature has been specifically requested by a large system OEM
(indeed, they have contributed already to the technical aspects as their
requirements have been refined).  Considering how we are always asking for
system (as opposed to vendor) input into the standards process, I personally
welcome their interest and hope that the standards process can accommodate
their requirements.  I would like the conference call to focus first on the
technical issues.  Time permitting, we can discuss the desirability of the
feature.  Honestly, I doubt that the issue will be resolved until the T10
meeting itself.

Finally, I believe that a motion will be made to start a SPI-4 document at
the next meeting.  Given the need to wind down SPI-3 activity, and our poor
history on SCSI-2 (it took 5 years to wind it down, mainly because we did
not have a SCSI-3 on deck), I'd like people to think of this as the first
SPI-4 feature (although if agreement is quick, and time permits, this could
be incorporated into SPI-3 - I'll defer to the editor on that topic).


PS  - I realize that at least one of you have a conflict with this time (as
I myself do - I had to reschedule a standing internal con call).
Unfortunately, this appeared to be the best time for everyone involved.

* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at

More information about the T10 mailing list