Draft Minutes of SPI-3 Working Group Meeting -- November 3, 1998

John Lohmeyer lohmeyer at ix.netcom.com
Fri Nov 13 13:46:08 PST 1998


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* John Lohmeyer <lohmeyer at ix.netcom.com>
*
At 11/12/98 07:14 PM, Gene_Milligan at notes.seagate.com wrote:
>The minutes failed to capture my request that if a proposal is not provided
>that  item 4.3 be removed from the SPI-3 agenda.

Gene, neither Ralph nor I recalled your statement -- which is why it did
not get recorded.  I assume by your above comment that you regard 98-156 as
being insufficient in some way.  I am hoping that Wally will be available
for the January '99 SPI-3 meeting.  If not, I think it appropriate to drop
this item until he (or someone) brings it up again.  I have added, "Gene
Milligan requested that this item be dropped from future SPI-3 agendas if a
new/revised proposal is not received."


><< Louis and Paul agreed to develop a mutually
>agreeable I-V curve for presentation at the next SPI-3 working group.>>
>
>     That is interesting. I heard them agree to defer further discussion
>until the next meeting. I did not hear them agree to agree.

Perhaps it is asking too much for competitors to agree, however, I think
Ralph recorded what was said correctly.  Nonetheless, I have deleted
'mutually agreeable' on the assumption that they may have been overly
optimistic.


>Regarding 4.8: <<Gene Milligan asked if there is a schedule for the
>development of SPI-3.  John
>Lohmeyer reported that the project proposal for SPI-3 says that forwarding
>to
>NCITS will occur in May, 1999.  Gene suggested that the group consider
>setting
>a cutoff date for new proposals.>>
>
>     The minutes are correct. But, I would like to point out the following
>items from T10's policies and procedures:
>     "It is the policy of X3T10 to: focus standards development within the

>defined scope of X3T10; control the development schedule to achieve timely
>publication; and to provide regular updates of adopted standards which need
>enhancements to fuel advancements in the industry.
>
>     The Chair will appoint  a project leader and a project editor for each
>new project for which the Chair has authorized a study effort and/or for
>which a project has been approved by X3. The project leader may or may not
>be the project editor. The project leader shall be responsible for
>providing a progress report including schedule adherence, key issues, and
>corrective actions at each plenary."
>
>     I request that the officers of T10 resume taking action to enforce
>these aspects of the T10 policies and procedures.

I believe that T10 is in compliance with its procedures.  If you know of
specific lapses, I would be glad to discuss them with you.


>Regarding 4.9: <<Gene Milligan noted that timing definitions now include QA
>maximum assertion
>and release times, but the defined times are not used in the protocol
>description. >>
>
>     I did not say "but the defined times are not used in the protocol
>description." I said "that as long as there is not a requirement that the
>assertion time be as long as to travel down a maximum cable delay to an
>expander, through the expander, and down a second maximum cable delay, the
>name changes had cleared the issue I had raised."

Sentenced changed to:  "Gene Milligan noted that timing definitions now
include QA maximum assertion and release times and that as long as there is
not a requirement that the assertion time be as long as to travel down a
maximum cable delay to an expander, through the expander, and down a second
maximum cable delay, the name changes had cleared the issue he had raised."


>I think 4.10 should have gone on to say that "there was general agreement
>at this session that the setup time should be approximately 10 nanoseconds
>longer than the data setup time."

Sentence added.

>Regarding <<4.14   QAS and Glitch Filters>> the minutes are fairly accurate
>except for the opinion that "but found no immediate solutions". Even that
>is correct if that means the suggested solutions were not immediately
>written down and were only spoken. I think the fact and statement that any
>glitch filters can be specified to be of an adequate width to ensure that
>the QAS message is seen by all QAS devices was missed by an impression that
>I think occurred that devices would be looking for the message in different
>data rate modes. There are no message rate modes for the message phase.
>Clearly the QAS message needs to be asserted long enough and the ACK/REQ
>needs to be long enough and other messages will need to have timing that

>does not produce a corner condition of false detection. Rather than making
>a rash proposal at the meeting it was more appropriate to consider Bruce's
>concern and to double check if the SPI-3 message timing requirements for
>QAS devices are OK or if they need changes.

I did not hear any 'spoken' solutions.  I did hear some suggestions that
some organizations thought they had approaches that would work.  I also
heard a suggestion that there could be interoperability concerns; i.e., one
vendor's solution may or may not work with another vendor's solution.

I am hoping that someone from the pro-QA camp will either offer a rebuttal
or a proposal on how to fix the SPI-3 wording to avoid the problem that
Bruce identified.  I suspect that there needs to be some limits established
regarding message transfer speeds and/or limits on the asynchronous glitch
filters employed while 'snooping'.

In any case, I think it fair to add the following statement to the minutes:
 "Some people present said that workable solutions exist."

Thanks for your comments!
John

--
John Lohmeyer                  Email: lohmeyer at ix.netcom.com
LSI Logic Corp.                Voice: +1-719-533-7560
4420 ArrowsWest Dr.              Fax: +1-719-533-7036
Colo Spgs, CO 80907              BBS: +1-719-533-7950

*
* 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