X3T10 SCSI WG Minutes 11/8-9/94

John Lohmeyer jlohmeye at ncr-mpd.FtCollinsCO.NCR.COM
Wed Nov 16 13:58:33 PST 1994


Accredited Standards Committee
X3, Information Processing Systems
                                        Doc. No.: X3T10/94-r0
                                            Date: Nov 14, 1994
                                         Project: 
                                       Ref. Doc.: 
                                        Reply to: J. Lohmeyer

To:         Membership of X3T10

From:       Weber/Lohmeyer

Subject:    Minutes of X3T10 SCSI Working Group  November 8-9, 1994

-----------------------------------------------------------------------------

                                    Agenda
1.    Opening Remarks

2.    Attendance and Membership

3.    Approval of Agenda

4.    Physical Topics
   4.1   SCSI-3 Parallel Interface - Low Voltage (SPI-LV) (94-201) [Ham,
         Lohmeyer]
   4.2   Review of SPI Rev 14 [Lamers]
   4.3   Review of SCSI-3 Fast-20 Working Draft Document [Lamers]
   4.4   Fast-20 Case Study [Harris]
   4.5   Fast-20 Data [Ham]
   4.6   3.3 Volt SCSI (94-164r1) [Aloisi]

5.    Protocol Topics

   5.1   SAM Mandatory vs. Optional Features [Monia]
   5.2   ACA Issues (94-189)
   5.3   System Recovery on 3rd Party Commands (94-184) [Houlder]
   5.4   Addressability of TARGET RESET task management function [Snively]

6.    Command Set Topics
   6.1   Proposed INQUIRY Command Enhancements (94-188r6) [Weber]
   6.2   Exception Handling Selection Mode Page (94-190) [Penokie]
   6.3   Partition Mode Pages for Tape (94-152) [Lappin]
   6.4   Command Extensions for PCMCIA (94-203) [Joslin]
   6.5   Multiple Port Operations (94-233) [Snively]
   6.6   What Next for SPC? [Weber]

7.    Other Topics
   7.1   German and Japanese Comments on CD 9316-1 (SCSI-2) [Lamers]
   7.2   Report on PFA (Predicting Failures) Study Group Meeting (94-219)
         [Penokie]
   7.3   Letter Ballot Comments Resolution on X3T10 Standards Development
         Policy & Procedure (94-198) [Milligan]
   7.4   Other Letter Ballot Results [Lohmeyer]

8.    Meeting Schedule

9.    Adjournment

-----------------------------------------------------------------------------

                              Results of Meeting

1.    Opening Remarks

John Lohmeyer, the Chair, called the meeting to order at 9:00 a.m., November
8, 1994.  He thanked Jeff Stai of Western Digital for hosting the meeting.

As is customary, the people attending introduced themselves.  A copy of the
attendance list was circulated for attendance and corrections.

It was stated that the meeting had been authorized by X3T10 and would be
conducted under the X3 rules.  Ad hoc meetings take no final actions, but
prepare recommendations for approval by the X3T10 Technical Committee.  The
voting rules for the meeting are those of the parent committee, X3T10.  These
rules are:  one vote per company; and any participating company member may
vote.

The minutes of this meeting will be posted to the SCSI BBS and the SCSI
Reflector and will be included in the next committee mailing.

2.    Attendance and Membership

Attendance at working group meetings does not count toward minimum attendance
requirements for X3T10 membership.  Working group meetings are open to any
person or company to attend and to express their opinion on the subjects
being discussed.

The following people attended the meeting:

         Name          S        Organization         Electronic Mail Address
---------------------- -- ------------------------- -------------------------
Mr. Richard Johnson    V
Mr. Rick Heidick       A  3M Company                usmmm54s at ibmmail.com
Mr. Norm Harris        P  Adaptec, Inc.             nharris at eng.adaptec.com
Mr. Lawrence J. Lamers A# Adaptec, Inc.             ljlamers at aol.com
Mr. Sassan Teymouri    V  Adaptec, Inc.             sassan at eng.adaptec.com
Mr. Neil T. Wanamaker  P  Amdahl Corp.              ntw20 at eng.amdahl.com
Mr. Charles Brill      P  AMP, Inc.                 cebrill at amp.com
Mr. Michael Wingard    A  Amphenol Interconnect
Mr. Ron Roberts        P  Apple Computer            rkroberts at aol.com
Mr. John Lohmeyer      P  AT&T/ NCR                 john.lohmeyer at ftcollinsco
                          Microelectronics          .ncr.com
Mr. Jason Albanus      P  Burr-Brown Corp.          Albanus_Jason at bbrown.com
Mr. Ian Morrell        P  Circuit Assembly Corp.    crctassmbl at aol.com
Mr. Joe Chen           P  Cirrus Logic Inc.         chen at cirrus.com
Mr. Louis Grantham     P  Dallas Semiconductor      grantham at dalsemi.com
Mr. Michael Smith      A  Dallas Semiconductor      msmith at dalsemi.com
Mr. Charles Monia      P  Digital Equipment Corp.   monia at starch.enet.dec.com
Mr. Ralph Weber        A# Digital Equipment Corp.   roweber at acm.com
Mr. William Dallas     A# Digital Equipment Corp.   dallas at wasted.enet.dec.com
Dr. William Ham        A# Digital Equipment Corp.   ham at subsys.enet.dec.com
Mr. I. Dal Allan       P  ENDL                      2501752 at mcimail.com
Mr. Kenneth J. Hallam  A  ENDL                      3450626 at mcimail.com
Mr. Edward Lappin      P  Exabyte Corp.             tedl at exabyte.com
Mr. Gary R. Stephens   P  FSI Consulting Services   6363897 at mcimail.com
Mr. Pat Edsall         O  Hewlett Packard Co.       edsall at hpdmd48.boi.hp.com
Ms. Nancy Cheng        O  Hitachi Computer Products n_cheng at hitachi.com
Mr. George Penokie     P  IBM Corp.                 gop at rchvmp3.vnet.ibm.com
Mr. Lansing Sloan      O  Lawrence Livermore Nat'l  ljsloan at llnl.gov
                          Lab
Mr. Kim Minuzzo        V  Lawrence Livermore Nat'l  minuzzo1 at llnl.gov
                          Lab
Mr. Dean Wallace       P  Linfinity Micro
Mr. Robert Bellino     P  Madison Cable Corp.
Mr. Pete McLean        P  Maxtor Corp.              pete_mclean at maxtor.com
Mr. Skip Jones         P  QLogic Corp.              sk_jones at qlc.com
Mr. Edward A. Gardner  A  Quantum Corp.             gardner at acm.com
Mr. Gene Milligan      P  Seagate Technology        Gene_Milligan at notes.seaga
                                                    te.com
Mr. Gerald Houlder     A  Seagate Technology        Gerry_Houlder at notes.seaga
                                                    te.com
Mr. Scott Smyers       P  Sony Corp. of America     scotts at lsi.sel.sony.com
Mr. Erich Oetting      P  Storage Technology Corp.  Erich_Oetting at Stortek.com
Mr. Roger Cummings     A  Storage Technology Corp.  Roger_Cummings at Stortek.com
Mr. Robert N. Snively  P  Sun Microsystems Computer bob.snively at eng.sun.com
                          Co
Mr. Vit Novak          A  Sun Microsystems, Inc.    vit.novak at sun.com
Mr. Bill Boyd          O  Texas Instruments         bboy%mimi at magic.itg.ti.com
Mr. Kevin Gingerick    V  Texas Instruments         4307725 at mcimail.com
Mr. Paul D. Aloisi     P  Unitrode Integrated       Aloisi at uicc.com
                          Circuits
Mr. Jeff Stai          P  Western Digital           stai at dt.wdc.com
                          Corporation
Mr. Tak Asami          A  Western Digital           asami at dt.wdc.com
                          Corporation

45 People Present

Status Key:  P    -  Principal
             A,A# -  Alternate
             O    -  Observer
             L    -  Liaison
             V    -  Visitor

3.    Approval of Agenda

The proposed agenda was approved with the following additions:

   4.4   Fast-20 Case Study [Harris]
   4.5   Fast-20 Data [Ham]
   4.6   3.3 Volt SCSI (94-164r1) [Aloisi]
   5.4   Addressability of TARGET RESET task management function [Snively]
   6.5   Multiple Port Operations (94-233) [Snively]
   6.6   What Next for SPC? [Weber]
   7.4   Other Letter Ballot Results [Lohmeyer]

4.    Physical Topics

4.1   SCSI-3 Parallel Interface - Low Voltage (SPI-LV) (94-201) [Ham,
      Lohmeyer]

John Lohmeyer described his project proposal for a parallel interface for
low-voltage environments (94-201r0).  Questions were raised regarding the
scope of the project.  John described the proposed document as a "delta
document" to SPI.  In John's scope, the document would not be a SPI-2.

Bill Ham asked if the study group, discussed at the last meeting, has taken
root.  John proposed that the study discussion occur immediately, at this
working group meeting.  John read the list of proposed items; battery-powered
and lower voltage power supplies, and enhanced cabling, connection, and
termination techniques.

Differences between low voltage and low power were discussed.  Proposals were
made for several different project proposals.  Counter arguments were raised
for a single, SPI-2 project proposal.

Paul Aloisi gave a formal presentation describing his thoughts about the
issues that should be covered in the proposed SPI-LV project.  Starting from
the issues already discussed, John led a group effort to produce a list of
next-generation SPI issues.  The list was as followed:

    1. Incremental and/or replacement standards documents
    2. Connectors (32-bit, ..., high-density)
    3. Higher speeds
    4. Lower voltage
    5. Lower power
    6. Hot swap (tm?) ... further definition
    7. System considerations
    8. Ease of use
    9. Cable specifications
   10. Backplane specifications
   11. Alternate topologies
   12. Enhanced error detection/correction
   13. Adaptive performance
   14. Fair access
   15. TERMPWR distribution
   16. Shield effectiveness
   17. Longer cable lengths

Duncan Penman asked the telling question, "Now what do we do on that list?" 
Gene Milligan suggested that more than two projects would result from the
list.  Folding existing work, like Fast-20, into the current SPI was
suggested and rejected.  Dal Alan asked Larry Lamers about the cost of
putting SCAM in SPI.  Larry said that the SCAM addition cost about 18 months. 
That is, SPI would have been a standard 18 months sooner if SCAM had not been
added.

Dal asked about market need and project difficulty for each item.  The
following table was produced where, the markets where divided into Niche and
Volume markets and these were further divided into Initiator (host) and
Target (peripheral) segments.  "H" means the feature is high priority, while
"L" means the feature is low priority for the indicated market segment.  The
"Stds dev" column indicates an estimate of the time and effort to complete
the standards documentation.

                                  ____Market____
                                  Niche   Volume
                                  I  T    I  T       Stds dev
 1. Types of Stds
 2. Connectors                    H  L    L  L       Long
 3. Higher speeds                 L  L    L  L       Medium
 4. Lower voltage                         H  H/L     Short
 5. Lower power                           H  L       Long
 6. Hot swap (tm?)                H  H               Short
 7. System considerations
 8. Ease of use                           H  H       Very Long
 9. Cable specifications          H  L               Short
10. Backplane specifications      H  L               Short
11. Alternate topologies          L  L               Long
12. Error detection/correction    L  L               Long
13. Adaptive performance          L  L    L+ L+      Short
14. Fair access                   L  L               Short
15. TERMPWR distribution          L  L    L  L       Long
16. Shield effectiveness          H  L               Medium
17. Longer cable lengths          H  L               Short

Having constructed the above list, the following subset seemed to be most
interesting (that is, high priority and/or short development time):

    4. Lower voltage [Aloisi]
    5. Lower power (power management) [Aloisi]
    6. Hot swap (tm?) ... further definition [Aloisi]
    8. Ease of use (list of topics) [Harris]
    9. Cable specifications (smaller) [Ham]
   10. Backplane specifications [Ham]
   13. Adaptive performance [Lohmeyer]
   14. Fair access [Penman]
   16. Shield effectiveness [Ham]
   17. Longer cable lengths [Ham]

Parenthetical comments generally represent additions to the descriptions that
were made while the responsible individuals were selected.  Names in square
brackets identify the people who signed up to bring proposals (or whatever)
to the next general working group meeting.

John said that additional work on the SPI-LV project proposal would be
deferred to the next working group meeting.  Bill Ham requested that item 2
(connectors) be remanded to SFF.  The people present agreed to Bill's
request.  

This briefly brought up the issue of bent connector pins, which had been
discussed on the SCSI Reflector.  John asked that anyone with actual examples
of bent pins contact their connector vendors for analysis; it appears that
only some organizations are experiencing these problems.  The connector
vendors are eager to help, but they cannot solve the problem without data.

4.2   Review of SPI Rev 14 [Lamers]

Larry Lamers announced that he has incorporated comments from Paul Aloisi and
John Lohmeyer.  Larry reported receiving no technical comments.  He believes
that SPI is done, except for some style and formatting issues.  Larry
believes that SPI is on track for forwarding to second public review at the
X3T10 plenary meeting on Thursday.

Gary Stephens said that he has comments that Larry has not seen.  Gary's
comments were discussed in a line-by-line review of SPI.  The first
controversial issue was a normative reference to SCSI-2.  The back reference
to the A-cable forces the reference to SCSI-2.  Other interesting issues were
claims that the B-cable is mentioned, driving of lines during BUS FREE phase,
confusion over de-glitching the RST signal, and concerns over whether a SCAM
initiator must be able to perform normal initiator functions.  All issues
were resolved with either minor wording changes or no wording changes.

4.3   Review of SCSI-3 Fast-20 Working Draft Document [Lamers]

Larry Lamers announced that Fast-20 is in about the same shape as SPI.  Larry
believes that a letter ballot could taken after the November Plenary for
forwarding Fast-20 to first public review.

Larry and Paul Aloisi discussed Paul's review comment regarding the location
of terminators in a Fast-20 configuration.  Paul objected to the note that
says, "SCSI devices shall not include termination."  Gene Milligan and John
Lohmeyer postulated that the note was added to keep the node capacitance
down.  Skip Jones noted that the Plug-and-Play SCSI industry specification
disallows any enabled terminators on devices.

John proposed that, if a switchable terminator is present, it shall
contribute to the node capacitance budget for the device.  Gene expressed
concern that any change should be made based on a plenary vote.  After some
attempts to create the wording, John proposed that the issue be taken to the
plenary with no recommendation from the working group.

Paul had a second concern regarding sinking regulators on active negation
circuits.  After some consideration, John suggested that the proposal will
outlaw some existing regulated terminators.  Again, John proposed that the
issue be taken to the plenary with no recommendation from the working group.

Kevin Gingerick questioned the transient leakage current requirement.  He
said that many devices will fail the test all the time.  Bill Ham suggested
dropping the requirement for the first 10 nanoseconds.

Kevin ask for an explanation for the 200 pF load in the single-ended test
circuit.  Bill Ham, John Lohmeyer, and Sassan Teymouri tried to explain the
reasoning behind the existing circuit.  Part of the concern was how well the
existing test circuit will model an actual system.  Kevin argued that no
capacitance is required to model a real system.  Bill and Sassan suggested
that around 100 pF would be appropriate.

After some discussion, the working group almost agreed on a maximum
capacitance of 210 pF.  A key part of the agreement was defining a maximum
capacitance.  Then, Sassan noted that use of maximum would generate
uncertainty in many of the timing measurements.  After further discussion,
the working group returned to the original wording, pending someone 
measuring total bus capacitance in some real-world configurations.

4.4   Fast-20 Case Study [Harris]

Norm Harris reported on studies of three Fast-20 bus configurations; a 3 m
4 load configuration, a 1.5 m 8 load configuration, and an external/internal
configuration with .75 m and 3 loads internal, and 1 m and 1 load external. 
The complete report can be found in document 94-232r0.

Norm's data compared assertion and negation times using various real-device
capacitance loads.  He also compared the affects produced by legacy and
Fast-20 specified cables.

Norm's major conclusion was that cable construction (particularly inside a
component box) are very significant in determining the limits to which Fast-
20 configurations can be taken.  In one example, Norm showed how poor
internal cable construction in a CD-ROM produced capacitance values that were
twice as bad as other equipment.  Bill Ham expressed concern that the
capacitance measurements might not be done as specified in the SPI.  Bill 
noted that the measurements must be biased as specified in SPI, or the
capacitance values might be much higher than they should be.  Paul Aloisi
showed some pertinent data regarding the effect of bias on capacitance
measurements.

Norm reminded the working group that his data suggests good head-room in the
Fast-20 specifications.  Norm is confident that the specified configurations
will work properly with Fast-20.

Next, Norm discussed his future plans.  Norm's plan to study mixed 8-bit and
16-bit devices took the working group down a major blackhole.  Bill described
problems with 16-bit lines passing through 8-bit devices.  Some lines have
capacitance load, but others do not.  Eventually, significant skew is
introduced between the loaded and non-loaded lines.

4.5   Fast-20 Data [Ham]

Bill Ham presented new data regarding Fast-20 bus length and device
capacitance.  The goal was to better understand the 3 meter bus length with
8 devices (7 clustered at the far end).  Bill discovered that to restore
'good' wave shapes, device capacitance must be lowered to somewhere between
3 pF and 10 pF.  Bill also noted that a point-to-point bus will work with
very good margins even at a length of 25 meters.  That is, bus loading
distribution makes a big difference.

Bill and Sassan Teymouri discussed cable construction effects on capacitance. 
Cables that are sliced at connection points (that is, the ribbon conductors
are separated back several inches) introduce much more capacitance loading
than cables that use insulation-displacement connectors.  Bill described
further configuration restrictions that improve the capacitance problems. 
For example, spacing the loads out evenly is much better than clustering
devices.

4.6   3.3 Volt SCSI (94-164r1) [Aloisi]

Paul Aloisi presented his revised proposal for changes to SPI that support
a 3.3 volt environment.  Bill Ham, Pete McLean, and Dal Alan discussed
TERMPWR tolerance limits, calling for different limits than Paul's document
suggested.

Paul asked about the format that his work should address.  Is the format for
inclusion in SPI-2 or for some smaller scope document?  No clear answer was
provided.

Paul noted that his proposed icon for systems that cannot provide TERMPWR has
been changed to a T overlaid with a circle-slash.

Paul discussed his proposal to allow TERMPWR to drop to 3.3 volts at the
source.  Bill noted that the proposal requires a regulated terminator.  Bill
also wanted to be sure that the 3.3 volts is the lowest limit at the source. 
Some of the proposal seemed to indicate that the source lower limit would be
3.0 volts.  Pete McLean said that JEDEC has accepted 3.3 volts plus or minus
0.3 volts.  Paul noted that his proposal has been applied to CD-ROM drives
connected to a laptop.  In that situation, Paul's proposal is working well.

Paul agreed to do more work on 3.3 volt SCSI for the January meetings.

5.    Protocol Topics

5.1   SAM Mandatory vs. Optional Features [Monia]

Charles Monia requested that the mandatory vs. optional features issue be
left open until after the public review of SAM.  Charles suggested that at
least one public review comment will be made on the subject.  John said that
the process does not guaranteed a public review comment just because there
is an outstanding letter ballot comment.  Gary Stephens seemed certain that
a public review comment will arrive.

5.2   ACA Issues (94-189)

Ralph Weber reported that 94-189r1 was approved at the September Plenary. 
Dal Alan reported problems with general understanding of ACA.  Dal stated
that the SSA and Fibre Channel interpretations of ACA are very different. 
Charles Monia took notes regarding the issues.  Dal asked for a four page
presentation on what ACA is expected to do.  Charles volunteered to make the
presentation.

After several people indicated that Bob Snively had problems with ACA, Bob
described his concern with an FCP disk profile that required Control<ACA> be
equal to one.  Bob described a meeting whose results were (in Bob's opinion)
that Control<ACA> is not necessary for well-constructed operating systems.

George agreed that Charles is the right person to make the four page
presentation on ACA, as requested by Dal.

5.3   System Recovery on 3rd Party Commands (94-184) [Houlder]

Gerry Houlder described how third-party reservations will not help with error
recovery on XOR-write commands.  The XOR study group will continue working
on this problem.  Gerry said that this problem may resurface after much more
study within the XOR study group.  However, Gerry made no predictions
regarding the ultimate results that might come from the XOR study group.

5.4   Addressability of TARGET RESET task management function [Snively]

Bob Snively raised concerns about TARGET RESET of targets buried inside
storage arrays (as per the SCSI-3 Controller Commands document).  Such
targets are identified by both a target identifier and a logical unit
identifier.  Charles Monia asked about the possibility of creating a LUN
RESET task management function.  Bob will revise his proposal based on
working group input.

6.    Command Set Topics

6.1   Proposed INQUIRY Command Enhancements (94-188r6) [Weber]

Ralph Weber reviewed changes made to 94-188 as a result of discussion on the
SCSI reflector.  Then Ralph discussed the three unresolved issues.

After some discussion, the working group recommended that the operation code
byte in the response data be changed from a constant FFh to the actual
operation code being reported.  A straw vote on this subject was 9 for
changing the returned data, 0 for keeping FFh, and 1 for removing the field
|from the returned data entirely.

Ralph asked for advice about what to do when the data is stored on media and
cannot be returned.  Currently, the Valid bit would be returned as zero. 
Someone on the SCSI reflector suggested returning CHECK CONDITION.  No one
at the working group defended or supported the use of CHECK CONDITION in this
case.

The last open issue concerned combining two CDB bits to form one 2-bit field
instead of two 1-bit fields.  Ralph and John Lohmeyer opposed the change
because it would produce confusion for people switching from SCSI-2 to SCSI-3
(while the bit would work the same, it would have a different name and would
not line up with SCSI-2 documentation).  Ed Gardner and Gerry Houlder noted
that the combined 2-bit field would be easier to explain.  Larry Lamers noted
that most currently shipping disks have vital product data pages,
particularly for serial numbers.  Thus, the SCSI-2 to SCSI-3 confusion issue
should be taken seriously.  Eventually, the no-change philosophy seemed to
prevail.

George Penokie asked that the proposal be extended to return a list of
supported operation codes.  Ralph suggested that George's proposal is a
separate issue, best introduced in a new proposal document.  After some
consideration, George agreed with Ralph.  George began describing the
proposal as a new vital product data page.

6.2   Exception Handling Selection Mode Page (94-190) [Penokie]

George Penokie described a long running problem (religious argument)
regarding reporting of asynchronous events, which include predicted failure
conditions.  94-190r2 is the current proposed solution for the problem. 
George reviewed the proposal in detail.  George noted that the scope of the
proposal has been limited to exceptions that report an additional sense code
of FAILURE PREDICTION THRESHOLD EXCEEDED.  Bill Dallas, Gary Stephens, and
Gerry Houlder provided substantial advice regarding improvements in the
proposal.  After extensive discussion, George promised another revision of
94-190.

6.3   Partition Mode Pages for Tape (94-152) [Lappin]

Ted Lappin described his proposed cleanup of the Medium partition page,
document 94-152r0.  Ted asked about the notes that he has added.  The notes
push the implementation of the standard in certain directions.  Ralph Weber
questioned whether the notes can be made normative.  Ted, Ralph, Bill Dallas,
and Gary Stephens discussed how best to handle the notes.

Gerry Houlder questioned the use of REFORMAT as a bit name.  Gerry was
concerned that the bit might be confused with physical media formatting.  Ted
agreed to change the bit name.  Bill Dallas asked that the proposal be
revised to clarify the additional sense codes reported for various error
conditions.

Ted agreed to discuss the proposal at the SSC working group meeting on
Thursday morning.  He also will revise the proposal.

6.4   Command Extensions for PCMCIA (94-203) [Joslin]

Philip Joslin was not present to discuss his document.  John Lohmeyer tried
to describe the proposal.  The hardware to be covered by the proposal is a
SCSI device that connects PCMCIA cards to the SCSI bus.  94-203 proposes that
block-ish PCMCIA devices be represented as block command SCSI LUNs.  However,
some changes are required in the SCSI-3 Block Commands.

In the absence of a presenter, the working group deferred action on this
matter.

6.5   Multiple Port Operations (94-233) [Snively]

Bob Snively described two problems in multi-port systems; overriding
reservations from a dead initiator and the cleanup of resources held by a
dead initiator.  Ralph Weber questioned how Bob's reservation override
concept relates to the PORT STATUS command, added to the SPC in revision 3. 
Bob said that he has not had time to study SPC r3, because it arrived in the
last two days.

Bob described his primary concept that all initiators are on all ports are
viewed equally.  Bob, Gerry Houlder, and Charles Monia discussed the
specifics of this concept.

Bob described the need to cleanup resources held by a known-to-be-dead
initiator.  The ABORT TASK SET, OTHER INITIATOR task management function is
proposed to resolve this.  Charles questioned the how knowledge of an
initiator identifier is available for use in the ABORT TASK SET, OTHER
INITIATOR function.  Ed Gardner and Scott Smyers suggested that the initiator
identifier can be passed among the initiators, using any of several
mechanisms that are beyond the scope of SCSI standards.

Gary Stephens noted that multiple SCSI Domains can be involved on both
parallel and Fibre Channel SCSI.  When that happens, initiator identifiers
might not be unique.  Bob described the solution, which is to simplify the
parallel case and to use world-wide names in Fibre Channel.

Bob asked the working group to review the proposal offline and provide
comments.  Bob promised to review the PORT STATUS command for applicability
to the problem.  Bob said that he expected that revisions to 94-233 will be
required before the document will be ready for inclusion in SCSI-3.

6.6   What Next for SPC? [Weber]

Ralph Weber asked the working group if the time had come for a letter ballot
on SPC.  There was some discussion of the outstanding proposals that affect
SPC.  Ralph noted that there are always outstanding proposals for SPC.  There
was no final resolution of the question.

7.    Other Topics

7.1   German and Japanese Comments on CD 9316-1 (SCSI-2) [Lamers]

Larry Lamers said that he has received over 200 comments from Japan and
Germany.  Of these, about 15 are comments that are not obviously correct. 
Larry already has made all the obvious corrections.  The working group
reviewed our positions regarding the comments that do not concern obvious
mistakes and/or clarifications in ANSI SCSI-2.

7.2   Report on PFA (Predicting Failures) Study Group Meeting (94-219)
      [Penokie]

George briefly reported on the results of the PFA (Predicting Failures) study
group meeting.  The minutes of the meeting can be found in document 94-219.

7.3   Letter Ballot Comments Resolution on X3T10 Standards Development Policy
      & Procedure (94-198) [Milligan]

On 94-208 (X3T10 Policies and Procedures), the letter ballot results were
47:7:1:3.  Yes-with-comments ballots were received from Cirrus Logic, Digital
Equipment, and Unitrode.  No votes were received from Adaptec, Amdahl, AT&T,
FSI, Oak Technology, Storage Technology, and Western Digital.

The attempt to resolve comments was lengthy, laborious, contentious, and
frequently descended to picking nits.  Time ran out before any of the
comments from Amdahl, AT&T, FSI, Oak Technology, Storage Technology, or
Western Digital could be discussed.  John Lohmeyer proposed forming a study
group to resolve the comments.  John noted that this would require action
|from the plenary.  The weary negotiators decided to move the effort to
tomorrow's plenary meeting (see item 10.1.5 in 94-227).

7.4   Other Letter Ballot Results [Lohmeyer]

John Lohmeyer reported the results of five letter ballots.  (N.B., all these
ballot results were revised several times during the week, as newly found
letter ballots were received and processed.  John blamed this problem partly
on his own FAX department, which had moved during the last two weeks.  See
94-231 for the final results.)

On 94-204 (Logging Operations TIB), the letter ballot results were 53:0:0:5. 
There were no comments received on 94-204.

On 94-205 (Sequential Access Partitions TIB), the letter ballot results were
53:0:0:5.  A yes-with-comments ballot was received from Seagate.  Ted Lappin
led a discussion of the comments from Gene Milligan (Seagate).  There were
some format issues that were deferred to Larry Lamers.  A comment about use
of "n" or "n+1" was rejected.  Agreement was reached regarding the resolution
of all comments.

The working group recommended specific resolution actions for all comments
to the plenary.  Note, all comments will be accepted except 6 (as noted
above).  Ted will revise the document based on the agreed changes as
discussed in the working group.  If the plenary approves, John will forward
the revised document for further processing.

On 94-206 (SCC forwarding to 1st public review), the letter ballot results
were 52:1:0:5.  Yes-with-comments ballots were received from IBM and
Unitrode.  The no vote came from Seagate.  George Penokie reported that the
SCC working group resolved all comments, except those received from Unitrode. 
George noted that all the Unitrode comments are editorial and that many of
them overlap the Seagate comments.

On 94-207 (SSA to SCSI-2 Mapping technical report project proposal), the
letter ballot results were 50:2:0:6.  The no votes came from Seagate and
Unitrode.  John reported that X3T10.1 has resolved the comments.

8.    Meeting Schedule

The next working group meetings will be the week of November 9-13, 1995 at
the Harrah's Lake Tahoe (702-786-3232) in Lake Tahoe, CA hosted by Silicon
Systems.  The room rates are $100.00 (plus 8% tax).  The reservation deadline
for these rates is December 1, 1994.  The group name is ANSI X3T10 or X3T10
Meetings.  The host contact is Steve Finch at: 714-573-6808, FAX: 714-573-
6916, or EMAIL: steve.finch at us.ssi1.com.

9.    Adjournment

The meeting was adjourned at 5:43 p.m. on Wednesday November 9, 1994.

-- 
John Lohmeyer                      E-Mail:  John.Lohmeyer at FtCollinsCO.NCR.COM
NCR Microelectronics                Voice:  719-573-3362
1635 Aeroplaza Dr.                    Fax:  719-574-5462
Colo Spgs, CO 80916              SCSI BBS:  719-574-0424 300--14400 baud




More information about the T10 mailing list