Draft Minutes of SPI-2/SPI-3 WG Meeting -- Mar 16, 1998

Gene_Milligan at notes.seagate.com Gene_Milligan at notes.seagate.com
Wed Mar 25 10:46:02 PST 1998


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* Gene_Milligan at notes.seagate.com
*
<<4.1     Staged Contact Resistance (98-xxx) [Herrman]>>

     Although the minutes do not mention it, the question was raised as to
whether there were any technical reasons the profiled overcoat could not be
applied instead to the host side of the connector. This would allow hot
plugging applications to selectively bear the added cost. I think the
question was not answered and should remain an open issue.

<<4.4     Bias Reduction Proposal (98-132) [Bridgewater]

...
It was noted that the proposal allows compensation for some of the fixed
tolerances currently present in the SPI-2 electrical signaling
requirements.
This led to a discussion of compensating for other tolerances by adding a
signal integrity testing procedure to the protocol.  Both ideas met with
general agreement, but issues around effects on higher protocol levels were
raised.  Wally agreed to bring a more detail proposal to the next meeting.
Wally then repeated his request that the open circuit LVD driver voltage be
reduced to 2.5 volts.  Noting that the request depends on a long succession
of
future performance improvements in parallel SCSI, some present suggested
that
no change may be necessary.>>
     It seems a large stretch to state that there was general agreement. It
is not even clear what both ideas were. It would be OK to state that there
was no violent objection at this point. It would be better to say nothing.
A document number is given but I do not recall Wally presenting anything
except what seemed to be a spontaneous chalk talk using blank overheads and
marking pens. It is also difficult to understand "some present suggested
that no change may be necessary" That might mean most thought a change was
necessary except for a few that were uncertain and thought a change may be
necessary."

<<4.5     Dual Clocking Proposal (97-208) [Bastiani]
Vince Bastiani reported that dual-edge signal testing is still in progress
and
requested that further discussion be deferred to the May agenda.
4.6     Test results on dual edge signals (98-113r0) [Bastiani]
Vince Bastiani reported that testing is still in progress and requested
that
further discussion be deferred to the May agenda.>>

     Other than there have been at least two papers on the topic, why are
there two agenda items for what appears to be the same topic? This comment
is directed more towards the May agenda than the March minutes.

<<an situation>>

     The secretary is definitely contaminated by "an SCSI".

<<4.7     Testing results on hot plugging perturbations (98-129r0) [Gintz]
...
Bill noted several glitches on all of the lines he sampled, as well as
concerns about transients caused by unplugging, an situation that he felt
the
group was ignoring.>>

     I think some disputed Bill's allegation of ignoring especially since
it is so early in the SPI-3 development. Probably the key point is not
characterizing the posture of the group but identifying whether or not the
enhancements implemented and those contemplated for hot plugging are
beneficial for both insertion and removal. It may be a semantics issue.
When I hear hot plugging I visualize both removal and insertion -
especially in fully populated systems. As I recall all our own hot plugging
testing has both elements.

<<4.9     QAS review (97-199r7 and 98-133r0) [Moore]
Richard Moore presented a overview (98-133r0) of latest revision of the
Quick
Arbitrate and Select (QAS) proposal (97-199r7).  In the course of the
discussion, it became clear to two expander vendors that the current
expanders
will not work with QAS.>>

     I think this is an accurate but incomplete rendering of what was said.
The omission makes the report less even handed. Current SCSI devices will
also not work in the sense of participating in QAS and consequently current
SCSI devices and current expanders will not be exposed to the current QAS
proposal. I think a more even handed report would be that "Two expander
vendors raised the concern that the installed base of their present design
would be confused by the way the proposal deals with the BSY line. Unless
the QAS proposal can be modified to accommodate such expanders, the use of
QAS would need to be limited at configuration time."

<<He expressed the belief that the next revision will not be the last.>>

     I am sure he did but we would be better served by leaving such remarks
out of the minutes. The only motive I can see for including such remarks is
to discourage people from expending effort to review the next revision.
Making the remark is fine, humorous, and relieves tension - that is not the
general purpose of the minutes.

<<4.10    CRC proposal for SPI-3 [Penokie]
George Penokie asked that discussion of this proposal be deferred to the
May
meeting.>>

     This is a confusing item. There is already a CRC proposal accepted
into SPI-3 for packetized. Somehow this agenda item should be labeled such
that it can be distinguished from the CRC already in SPI-3.








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





More information about the T10 mailing list