Agenda -- X3T9.2 Meeting #88 June 18-19, 1990 -- Wichita, KS 1. Opening Remarks 2. Approval of Agenda 3. Attendance and Membership 3.1 Roll Call of Members in Jeopardy 4. Approval of Minutes - April Meeting St. Petersburg, FL (X3T9.2/90-066) 5. Document Distribution 6. Liaison Reports 6.1 ISO [Milligan] 6.2 IT8 [Lohmeyer] 6.3 SCSI-2 Common Access Method Committee Report [Allan] 6.4 Fiber Channel [Allan] 6.4.1 Low-Cost Transceiver Modules [Kubinec] 6.5 X3B7.1 Report 6.6 Small Form Factor Meeting 7. Review of Old Action Items 8. Working Group and/or Project Status Reports 8.1 ESDI [Allan] 8.2 SCSI General Working Group (X3T9.2/90-76) 8.3 Cable SPecification Working Group Meeting (X3T9.2/90-78) [Lamers] 9. Old Business 9.1 Single-cable 16-bit SCSI (X3T9.2/90-48R1) [Tuesday Morning] 9.1.1 Cable Skew Testing 9.2 Packetized SCSI (X3T9.2/89-130R1) [Stephens] 9.3 Multi-Ported SCSI (X3T9.2/89-133R1) [Stephens] 9.4 Additional SCSI Caching Control (X3T9.2/90-021R1) [Milligan] 9.5 SCSI-3 Document Structure 9.6 Initiator Control of Reconnection (90-62R2, 90-83R0) [Lohmeyer] 9.7 RESERVE Command Conflict with Direct-Access Model (90-59, 90-85) [Penokie] 9.8 Interaction on LOG SELECT TSD and DS bits (90-067) [Milligan] 10. New Business 10.1 Write Protect Editorial Error (90-70) [Snively] 10.2 Reduced Common-Mode Range on Differential SCSI [Jessen] 10.3 ESDI Guidance and Interpretations (90-74R1) [Lamers] 10.4 Review of new documents 10.5 Agenda for the Rochester Working Group 10.6 SCSI-2 Alt-2 Termination for the B Cable [Wicklund] 11. Review of Action Items 12. Meeting Schedule 12.1 General Working Group Schedule for 1990 12.2 Financing Plenary Meetings 13. Adjournment Minutes -- X3T9.2 Meeting #88 June 18-19, 1990 -- Wichita, KS 1. Opening Remarks John Lohmeyer, the chairman, called the meeting to order at 9:00 a.m., Monday June 18 23, 1990. He thanked the committee members for attending the meeting in Wichita which allowed him to be home for Father's Day. He also thanked Classic Conferences for arranging the meeting. John noted that this was the last meeting arranged by Classic Conferences and that committee members should check the meeting minutes in their mailings for the future hosts. John pointed out that two working group meetings were scheduled on Tuesday with the first starting at 10:00 a.m. He asked that the members to help in expediting the meeting so that the plenary meeting could finish by that time. As is customary, the people attending introduced themselves. A copy of the X3T9.2 membership list was circulated for attendance and corrections. Copies of the draft agenda and the recent document register were made available to those attending. Information on X3T9.2 and Mailing Subscription Forms were made available. 2. Approval of Agenda The draft agenda was approved with the following additions: 6.6 Small Form Factor Meeting 10.6 SCSI-2 Alt-2 Termination for the B Cable [Wicklund] 3. Attendance and Membership The membership requirements were reviewed. X3 rules permit one vote per organization. An individual from a new organization must attend two consecutive plenary meetings. He/she may apply for voting membership the second meeting via a letter to the chairman of X3T9.2. The individual may vote at the second plenary meeting. Working group meetings do not count toward attendance requirements. The chairman indicated that since the last meeting, the following membership change had occurred: Archive has bought Cipher Data Products and Cipher/Optimem. Ed Young and Adrienne Turenne are the new principal and alternate members of the combined company replacing Bharat Shah and Dan Davies. Conner Peripherals became a member at the meeting with Steve Cornaby the principal member and Paul Thompson the alternate member. The membership of Standard Microsystems Corp. was reinstated at the meeting with Michael Britch the principal member and Richard Nesin the alternate member. With the addition of Conner and Standard Microsystems at the meeting and the termination of Burndy Corp. for non-attendance (see 3.1 below) X3T9.2 voting membership is now at 68 up from 67 organizations. Enclosure (1) is the list of attendees at the meeting. Enclosure (2) is the X3T9.2 membership changes since the last minutes and enclosure (3) is the current X3T9.2 membership list. 3.1 Roll Call of Members in Jeopardy The chairman said he had sent 15 jeopardy letters to 6 organizations: Ancot Corp., Apple Computer, Burndy Corp., Datacopy Corp., Exabyte Corp. and Northern Telecom, Inc. The letters said that these organizations had failed to attend two of the last three plenary meetings and that their membership would be terminated unless they attended the June plenary meeting. The letters also informed these people of the committee's ability to retain their membership under special circumstances in spite of their failure to attend. John said that he had received a letter from Jerry Walker, VP of Engineering at Exabyte, saying that Jim Rochester had attended the April meeting and did not show up in the attendance data. Consequently, Exabyte is not in jeopardy at the June meeting. Members were present from all of the organizations in jeopardy except Burndy Corp. The Burndy principal and alternate were changed to observer status. 4. Approval of Minutes - April Meeting St. Petersburg, FL (X3T9.2/90-066) The minutes were approved with two changes. The omitted list of new documents at the April meeting will be added and the note about Dal's editing changes will be deleted. Revision 1 will be posted on the SCSI BBS, but not re-distributed in the committee mailing. 5. Document Distribution The chairman noted that the mailing had been delivered to him three weeks ahead of the meeting. He did not receive any comments concerning the mailing, except that one person nearly missed it due to its small size (136 pages). Members who have any problems with their mailing subscription should deal directly with Katrina at (202) 626-5741. John Lohmeyer requested that all documents for the July mailing be sent to him by July 16, 1990. The following new or revised documents were distributed and/or discussed at the meeting: Document Doc Date Author Description of Document ------------- -------- --------------- --------------------------------------- X3T9.2/90-83 6/8/90 G. Houlder Alternative Initiator Control of Reselection X3T9.2/90-84 5/29/90 T. Wicklund Comments on Maxtor questions about ESDI (ref:X3T9.2/90-074R1) X3T9.2/90-85 6/7/90 J. Lohmeyer Proposed clarifications to SCSI-2 re Reservations X3T9.2/90-86 6/8/90 T. Wicklund SCSI-2 Caching Page Problem X3T9.2/90-87 6/12/90 T. Wicklund Sequential Device Partition Definition X3T9.2/90-88 6/14/90 T. Wicklund Caching Page Proposal (Document 90-021 R2) X3T9.2/90-89 6/15/90 B. Snively Proposed requirement for Diagnostic Command Set X3T9.2/90-91 6/18/90 J. Fiala Crosstalk/Capacitance Testing X3T9.2/90-92 6/17/90 E. Jessen Alternative Differential Standard and its impact X3T9.2/90-93 6/18/90 K. Chan Single-ended crosstalk measurement results X3T9.2/90-94 6/14/90 J. Newman Comments on Maxtor proposal for ANSI ESDI Specification X3T9.2/90-95 6/18/90 J. Kubinec Serialized Data Communications Presentation The current document register is enclosure (4). 6. Liaison Reports 6.1 ISO [Milligan] Gene Milligan provided the following report: ISO REPORT A SC 25 US Technical Advisory Group (TAG) meeting was held in mid May. The recommended US positions resulting from the April X3T9 meeting were agreed to. Judy Anderson, of EIA, was able to make arrangements accommodating the January SC 25 and related working groups. The IR acquired an electronic copy of the ANSI SMD-E X3.91M-1988 text file for the purpose of incorporating the ISO editorial comments and preparing DIS 9324 for publication. The tape SCTD X3.146-1987 is continuing to encounter delays in conversion to IS 9317 for publication. The latest and presumably last delay was due to the original company responsible for editing of the ANSI document declining to support the revision of one of the figures. The editor Lou Domshy is attempting to obtain a low key solution to redrawing the figure. IPI documents Physical IS 9318-1, IPI-2 Disk IS 9318-2, IPI-3 Disk IS 9318-3, and IPI-3 Tape DIS 9318-4 are close to editorial completion based upon the review of Brannon's comments on the IPI-3 Tape. Roger Cummings invited Brannon to the Wichita meeting as he was to be in the United States this week. Keith is expected Thursday. The IPI-Enhanced Physical DP 9318-6 editorial effort resumed after the April decisions on IPI editing. X3T9 should make a recommendation on when it should be submitted. The IPI-3 COM DP 9318-7, is still hung up due to the lack of an ISO editor. X3T9 should make a recommendation this week to break the deadlock. Although the IPI-2 TAPE DP 9318-5 letter ballot closed earlier this year, no results are yet available. The IR checked with the SC 25 Secretariat and was advised that the ISO Central Secretariat has not provided a tally. The IR is attempting to reach them in Geneva. The revision of ESDI DIS 10222 is still in process. The IR forwarded the recommended US yes vote on SCSI-2 DP 10288 with a recommendation to substitute Revision 10C for the DIS basis. The results of letter ballots on the HIPPI DP letter ballot and NWI letter ballot are pending. The IR forwarded a US recommended vote on the FDDI HRC DP letter ballot. The recommendation was for a No vote with recommended resolution per the revised ANSI document. The IR presented the FDDI-SMFPMD NWI to the SC 25 TAG. 6.2 IT8 [Lohmeyer] John Lohmeyer said he had some information concerning IT8's new project which he would make available upon request. 6.3 SCSI-2 Common Access Method Committee Report [Allan] Dal Allan reported. The ATA meeting on May 30 is likely to be the last one which will be held under the auspices of the CAM Committee. Future work on ATA will be done under X3T9.2. The outstanding technical item on whether the protocol for PDIAG and DASP should be changed was decided in favor of no change. The ATA project proposal is at X3T9 and is expected to be forwarded to X3 on Friday. CAM members are in the process of a letter ballot on ATA Rev 2.1. Microsoft hosted the May 31-June 1 CAM Committee meeting at their campus in Bellevue, with the best audio-visual facilities ever seen at a CAM meeting. Although the approach taken is similar to CAM, it was difficult to draw direct comparisons because the acronyms used were different and detailed specifications were not yet available. The proposal for OS/2 was described under the name of WAMIC, and has since been formally announced as LADDR. Criticism of the overview presented may have been affected by the lunch provided - the CAM members were unusually mellow after being served BBQ and beer. Microsoft is in the process of implementation, with a tight schedule for early release of their SCSI support. WAMIC was developed in conjunction with several companies. Documentation will not be available through the CAM Committee but directly from Microsoft on request, because they are making preliminary software drivers available, and need to know who to send floppies to. A signup sheet is included in the current CAM mailing. The WAMIC proposal is public domain, there will be no license fees (though you will have to pay for the implemented code modules if you buy a WAMIC implementation), and Microsoft wants to see it become an ANSI standard. A special working group had been set up for June 14 to review the detailed documentation but a late mailing prevented that. T.S. Lee of DTC, is in charge of arrangements for a meeting to be held on June 21, which either will be hosted by Syquest in Fremont or will be in Bellevue again. Whether the XPT/SIM and WAMIC will be merged into a single document or WAMIC become a stand-alone definition for OS/2 has not been decided, and will be based upon the review of the detailed documentation. Nothing was presented on DOS or Novell, and it seems that Microsoft will be following the IBM implementation of Interrupt 4Bh for DOS. IBM made a presentation on Subsystem Control Block Architecture on the second day, and will provide documentation to those who request it. The meeting wrapped up with a review of the present XPT/SIM document, which was updated to Rev 1.8. John Lohmeyer said he would post the newest revisions of both the ATA and XPT/SIM documents on the SCSI Bulletin Board System when he receives electronic copies. 6.4 Fiber Channel [Allan] Dal Allan reported. The working group is mostly discussing FC-2 right now, with an active FC-0 special working group operating in parallel. There is a strong likelihood of being able to source inexpensive fiber components soon (within 2 years), as indicated by AMD when the FOXI module was announced. IBM Endicott brought along a 25 MBs laser transmitter/receiver package which can operate at up to 10 km. The special working group dealing with the fabric issues held its final meeting in Toronto. Future fabric issues are to be dealt with in the regular working group activities on FC-2. The FC-3 application interfaces for SCSI, HIPPI, IPI-3, and Block Multiplexer Channel (FIPS-60) have been promoted to FC-4, and now FC-3 will be defined as common services for the using interfaces. More participation in the Fiber Channel group is needed to ensure that SCSI needs are protected in the FC-2 and FC-3 protocols. 6.4.1 Low-Cost Transceiver Modules [Kubinec] (90-095) Jim Kubinec of AMD gave a presentation on Serialized Data Communications (see 90-095). AMD has a current product called the TAXI chip which provides 125 Mbaud serialization, deserialization, coding, and decoding functions. They have married the TAXI chip with 1300 nm optical transmitters and receivers to make assemblies that are called FOXI modules. The transfer rate is 125 Mbaud/sec with 20% overhead so it can transfer data at 100 Mbit/sec. Current packaging is an assembly of the chip plus optical component in a 28-pin DIP. Plastic injection molding could be used to reduce costs. Current pricing is $295 per pair in volumes of 1000. With high volumes, there is a potential to get the pricing down to $25 per pair. 6.5 X3B7.1 Report John Lohmeyer stated that the minutes of the March 27, 1990 X3B7.1 committee meeting confirm that X3B7 will disband at the end of the year. Frank Meijers, the X3B7 chairman, has resigned due to pressing business commitments. George Canevit has replaced Frank as chairman of X3B7.1. The X3B7.1 members hope to publish their papers through IDEMA (International Disk Equipment Manufacturers Association) for the DISCON show. John also announced that X3B7.1 planned to meet next June 26-27, 1990 in Phoenix at The Pointe at Tapatio Cliffs (602-866-7500). 6.6 Small Form Factor Meeting The Small Form Factor (SFF) meeting was held May 29-30 to discuss mechanical and electrical issues related to sub-3.5 inch disk drives. The catalyst for this meeting was one held by Sun Microsystems to learn what vendors were doing about interoperability in small drives. The results were somewhat discouraging and Dal Allan was asked to form an industry consortium on similar principles to the CAM effort which he is chairing. An estimated 70 people attended. The primary focus of the discussion was the 2.5-inch form-factor drives. There was little support for standardizing anything on the 2.5-inch ATA interfaced drives. Most vendors planned to follow the Conner lead for both ATA and SCSI. Sun made a strong effort to achieve some consensus on SCSI drives in the 2.5-inch form-factor. Bob Snively met with several people on Tuesday evening and came in with a proposal on Wednesday. This proposal was modified during the discussion Wednesday, but there seems some hope of having a mounting hole agreement, and a signal pin-out agreement. Basically, Sun defined an environment for small drives which was different to that first envisaged for this form factor. The meeting ended without a decision on what will happen next and no future meetings are scheduled unless at least 10 companies are willing to actively participate as members in an industry effort. Sun Microsystems plans to distribute the results of the first meeting to all who attended. 7. Review of Old Action Items 1. John Lohmeyer will write a letter requesting SMC clarify the termination procedures and provide guidelines regarding termination exceptions. Carry over. 2. John Lohmeyer will request that X3T9 recommend the US position on the ISO/IEC draft proposal letter ballot on SCSI-2 be that the document should be replaced with the SCSI-2 Revision 10c document. Complete. 3. John Lohmeyer will request that X3T9 submit SCSI-2 Rev 10c to X3 for the X3 letter ballot. Complete. 4. John Lohmeyer will submit the Common Access Method Project Proposal (X3T9.2/90-57) to X3T9. Complete. X3T9 changed the name of the Project Proposal to "SCSI Common Access Method" 5. John Lohmeyer will submit the AT Attachment Project Proposal (X3T9.2/90-58) to X3T9. Complete. 8. Working Group and/or Project Status Reports 8.1 ESDI [Allan] A meeting was held June 12, 1990 to "discuss possible future enhancements to the ESDI standard including zoned recording, alternative connectors, and faster transfer rates." Dal reported that Said Zengenhpour of Zenith had been responsible for setting up this meeting as a way to increase the awareness of drive and controller manufacturers that at least one company wanted to pursue the enhancement of ESDI and the implementation of notched disks as defined in the standard. About 50 people attended, several of whom were from connector companies. The connector company representatives left as soon as it was determined that there would be no new connectors defined. The conclusion of the meeting was that enhancing the present physical interface which had been based on ST506 was a major problem and that it might be better to port ESDI over to a serialized implementation based on the Fiber Channel (using copper cabling). The support for ESDI drives is a market issue that does not need any new standards activity. 8.2 SCSI General Working Group (X3T9.2/90-76) A working group meeting was held May 8-9, 1990 at Providence, RI with 25 people present. The group did not let a short agenda stop them from having a long meeting. The working group discussed several areas in the SCSI-2 draft standard that should be clarified during the final editing. Motions authorizing these clarifications were planned for the Wichita plenary meeting. 8.3 Cable Specification Working Group Meeting (X3T9.2/90-78) [Lamers] The Cable Specification Working Group met May 7, 1990 in Providence, RI with 27 people present. The group put together a spreadsheet compiling the results of the various cable tests. An attempt was made to correlate the parameters with the three cables that Sun Microsystems identified as passing their system tests. The attempt failed partly because the crosstalk testing had not been done in a manner that guaranteed adjacent pairs were tested. Rather than repeat all the testing, Jim Fiala (3M) volunteered to verify that capacitance testing would correlate to crosstalk. If so, capacitance testing could be done much more easily. A follow-up meeting was scheduled for June 19 beginning at 1:30 p.m. [Note: See X3T9.2/90-99 for minutes of this meeting.] 9. Old Business 9.1 Single-cable 16-bit SCSI (X3T9.2/90-48R1) [Tuesday Morning] George Penokie had nothing new to discuss on this topic. The proposal is ready to be incorporated into the SCSI-3 physical draft document, when it is prepared. 9.1.1 Cable Skew Testing At the last meeting it was recognized that there might be skew problems when mixing 8-bit and 16-bit devices. Tom Wicklund of Ciprico stated that he might be able to supply some devices to assist testing this. This issue was raised by chip vendors. Mismatched cable could cause stair steps in the rise and fall of signals. If the cable impedances are correct, there should not be a problem, but it would be great to verify there is no problem. 9.2 Packetized SCSI (X3T9.2/89-130R1) [Stephens] There nothing new on this topic. John urged people to read this document because it is a necessary step to move toward using the Fiber Channel interface. 9.3 Multi-Ported SCSI (X3T9.2/89-133R1) [Stephens] The Providence working group spent considerable time discussing multi- ported SCSI and did not reach a consensus over whether a simple dual- ported SCSI should be defined or a full multi-pathing SCSI as in Gary's 89-133R1 document. Gary's proposal allows multiple ports which may be grouped, that is, the ports are given identifiers. Ports with the same identifier may be used interchangeably for each connection of an I/O process. While this proposal allows great flexibility and robustness, it also requires that I/O processes be identified uniquely across the multiple ports. This is a new concept for SCSI which will require some kind of an I/O process identifier field. 9.4 Additional SCSI Caching Control (X3T9.2/90-021R2) [Milligan] Gene Milligan presented revision 2 of his proposal which contains the recommendations from the working group. He spoke against most of these. In particular, he prefers to include the control over usage of the MODIFY DATA POINTER message in the caching page rather than in the Control Mode Page. There was a discussion over why control of the MODIFY DATA POINTER (MDP) message needs to be separate from the Force Sequential Write (FSW) bit. While it is unlikely that the MDP message would be used when FSW bit is one, it is useful to provide both states of MDP control when the FSW bit is zero. The Forced Sequential Read (FSR) bit appears to mostly be useful as a diagnostic and performance measurement function. Gene will reconsider whether it should be included. Gene plans to revise the document in hopes of getting less contention. 9.5 SCSI-3 Document Structure At the Providence working group there emerged a consensus on developing a SCSI-3 physical standard that would deal with the P-cable and multiple porting. John Lohmeyer accepted an action item to draft a project proposal for this effort. [Note: The approved SCSI-3 project proposal assumes a single large SCSI-3 document. Since we are now planning to split the document into separate standards, a new project proposal is needed for each standard.] 9.6 Initiator Control of Reconnection (90-62R2, 90-83R0) [Lohmeyer] Since NCR had a vested interest in this issue, John Lohmeyer asked Dal to chair this portion of the meeting. Tom Wicklund gave a brief overview of the special needs for array applications using the presentation materials he prepared for the last working group. In a disk array, synchronizing the reselection process is necessary to optimize performance. John Lohmeyer gave an overview of his proposal and described how it had evolved over the last several meetings. Gerry Houlder presented an alternative proposal that put the function of the messages into the control byte of the CDB. John argued that if the TDRD bit is put in the CDB it should not be in the control byte since only the READ and WRITE commands need this control. John Lohmeyer moved and Tom Wicklund seconded that document 90-62R2 be accepted as the method of providing initiator control of reselection order for SCSI-3. Gene Milligan moved and Steve Cornaby seconded that the motion be amended to replace the TDRD message with a bit in the control byte as proposed in 90-83. The amendment failed: 9 yes and 14 no. A friendly amendment to incorporate the last two paragraphs of 90-83 (deleting "IMPLEMENTORS NOTE") was accepted by John and Tom. These paragraphs enable the initiator to reject the DISCONNECT message during a target reconnection in order to transfer data during that connection. The main motion passed: 24 yes and 3 no. A question was asked whether the TDRD message could be used on a subsequent connection to block data transfer on that connection. This issue was deferred to the Rochester working group. 9.7 RESERVE Command Conflict with Direct-Access Model (90-59, 90-85) [Penokie] There was considerable discussion over this issue at the last plenary meeting and at the working group meeting. The result of the working group investigation was that the intention of the Costa Mesa 1989 working group was that only the "allow" form (Prevent=0) of this command be permitted through a reservation. John Lohmeyer prepared document 90-85 to reflect what he thought was the consensus on several editorial clarifications related to the RESERVE and RELEASE commands in sections 8 and 9. John Lohmeyer moved and George Penokie seconded that document 90-85 be accepted as a non-substantive editorial clarification to resolve the RESERVE command wording conflicts. The motion passed: 28 yes and 1 no. The "no" vote was from Paul Jackson of IOMEGA. Since Paul had not objected during the discussion of the motion, Dal asked him why he had voted no. Paul indicated that he preferred allowing both variations of the PREVENT ALLOW MEDIUM REMOVAL command through a reservation. Dal explained that while it would make checking for reservation conflicts a little easier, such a capability would have little benefit since the initiator can not read the medium to determine whether it is the desired volume; READ commands are rejected while a reservation is active. Dal offered two alternatives 1) leave the document as is, or 2) change the document for editorial consistency. A straw poll indicated that only two people favored leaving the ambiguities in the document. 9.8 Interaction on LOG SELECT TSD and DS bits (90-067) [Milligan] The working group recommended deleting some confusing text about setting both the DS and the TSD bits in the log parameter page. Gene Milligan moved and Larry Lamers seconded that the last two sentences of paragraph 13 of section 7.3.2 be deleted as a non-substantive editorial clarification to resolve a conflict in log parameters page. The motion passed: 22 yes and 0 no. The Technical Editor was directed to delete the sentences during the final editing. 10. New Business 10.1 Write Protect Editorial Error (90-70) [Snively] Bob Snively's 90-70 document points out an inconsistency in how the Write Protect bit is described in several sections. In particular, section 8 fails to describe the MODE SELECT usage of this bit at all. Larry Lamers accepted a working group action item to draft consistent wording and to propose it at the plenary meeting. Larry Lamers presented a foil with consistent wording and its affect of each of the sections. This wording states that the WP bit is not defined for MODE SELECT commands. Larry Lamers moved and Gene Milligan seconded that the proposed wording be accepted to clarify the definition of the Write Protect bit. The motion passed 24 yes and 0 no. The Technical Editor was directed to incorporate the change during the final editing. While Larry was preparing the foil for the WP bit, he also noticed a slight inconsistency in the descriptions of the DPOFUA bit. He prepared a second foil describing his proposed editorial improvement to make these descriptions consistent. Bill Spence moved and Gene Milligan seconded that the proposed wording be accepted to clarify the definition of the DPOFUA bit. The motion passed 22 yes and 0 no. The Technical Editor was directed to make the change during the final editing. 10.2 Reduced Common-Mode Range on Differential SCSI (90-092) [Jessen] Erik Jessen proposed reducing the common-mode voltage range on SCSI differential drivers and receivers. This would allow a single controller chip to be designed with on-board differential transceivers. This new chip would cost around $16 and dissipate on the order of 2 watts. If termination doesn't change and no ground shift occurs this part could be compatible with current differential devices. Gary Murdock (National Semiconductor) expressed concern about AC ground shifts due to cable inductance; he agreed that there is not likely to be much DC ground shifting. Further testing is needed. The topic was remanded to the working group. The marketing question of whether further differential improvements should be considered if they meant that new products would not be compatible with old products was briefly explored. No consensus was reached since no one was ready to quantify the improvements. Consensus was reached on the point that the current differential specification would benefit from further clarification. 10.3 ESDI Guidance and Interpretations (90-74R1) [Lamers] Dal Allan, Larry Lamers, Gene Milligan, and Tom Wicklund met Monday evening to discuss what should be done with the submittal from Maxtor and the Hitachi response which Dal had received late Friday afternoon. Their recommendation was that no technical changes should be made to the ESDI document, but some editorial clarifications should be made which correct the ambiguities pointed out in the above documents. The expanded Vendor Unique status reporting proposed by Maxtor can be done in a vendor unique manner (which is permitted by the standard). 10.4 Review of new documents 90-80 [Spence] Proposal for Note re Aborting Arbitration The request for a note arose from a real problem where a device started arbitration and won, but failed to assert SEL (the device "changed its mind" about wanting the bus and simply dropped out of arbitration). Larry Lamers maintained that the current SCSI-2 wording prohibits this deviant behavior because the device failed to assert the SEL signal as required in step (4). He argued against adding Bill's proposed implementors note because it might be construed as permitting the unwanted behavior. A compromise to aid comprehension was reached whereby the current implementors note (in 5.1.2, paragraph 7) will be expanded by adding the following wording: "(1) The above procedure requires that an SCSI device complete arbitration once it starts the procedure until the point where the SEL signal is asserted." 90-86 [Wicklund] SCSI-2 Caching Page Problem Tom's document points out that there is an undefined state when the pre- fetch is more than the minimum but less than the maximum amount. Tom's document would clarify what to do in this case. Gene Milligan objected to the word "specifies", looking for a blander word such as "indicates". Tom Wicklund moved and John Lohmeyer seconded that 90-086r0 be accepted as an editorial clarification to SCSI-2 with the substitution of "indicates" for "specifies". The motion passed: 15 yes, 0 no, and 11 abstained. 90-87 [Wicklund] Sequential Device Partition Definition John Lohmeyer said he had received a FAX from Brian Earle of Eventide on the same topic, the interpretation of the partition page. Late in the meeting, John received Brian's formal request for interpretation on this topic. John will place it on the August plenary agenda and it was assigned to the working group. Bharat Shah agreed to investigate the issue. A second issue that Brian has requested interpretation on is whether filemarks and setmarks should receive a block address or not. The current working in the SCSI-2 draft standard allows either interpretation and would create a problem read a piece of medium using an absolute block number on drives from different vendors. Reading the the same block number would return different blocks because intervening filemarks or setmarks might be counted in one implementation and not counted in the other implementation. This issue will also be placed on the August plenary agenda and was assigned to the working group for investigation. 90-89 [Snively] Proposed requirement for Diagnostic Command Set Since Bob was not present to introduce his document, it was assigned to the July working group where this topic will be discussed on Wednesday morning. 10.5 Agenda for the Rochester Working Group The following agenda items were identified for the Rochester, MN working group meeting July 10-11, 1990: 1. Caching Proposal (90-021R2) [Milligan] 2. Diagnostic Command Set (90-022) [Pickford] {Wednesday a.m.} 3. Proposed requirement for Diagnostic Command Set (90-89) [Snively] 4. Packetized SCSI (89-130R1) [Stephens] 5. Multi-ported SCSI (89-133R1) [Stephens] 6. 16/32-bit cable issues (90-48R1) [Penokie] 7. Can TDRD be used to block data transfer on current connection? 8. SCSI-2 Request for Interpretation #1 (90-97) [Earle] {Should there be a Page Size Descriptor for each partition on the tape?} 9. SCSI-2 Request for Interpretation #2 (90-98) [Earle] {Should filemarks/setmarks be assigned block addresses?} 10. Reduced Common-Mode Range on Differential SCSI (90-092) [Jessen] 11. SCSI-2 Alt-2 Termination for the B Cable [Wicklund] 12. SCSI-3 Physical Project Proposal [Lohmeyer] 10.6 SCSI-2 Alt-2 Termination for the B Cable [Wicklund] Tom Wicklund pointed out that the Alternative 2 termination is specific to 8-bit SCSI on the A Cable. In particular, Figure 4-9B only shows the A Cable signals. John suggested that the figure be made generic by showing "-Signal" on one resistor, instead of showing the 18 A-cable signal names. He pointed out that more changes are probably also necessary because the B cable has more signal lines than the A cable and most likely needs a higher Imax specification on the voltage regulator. Tom Wicklund said he would analyze this issue further. 11. Review of Action Items 1. John Lohmeyer will write a letter requesting SMC clarify the termination procedures and provide guidelines regarding termination exceptions. 2. John Lohmeyer will draft a project proposal for SCSI-3 Physical. 3. Larry Lamers will incorporate the reserve/release, the write protect bit, and the DPOFUA bit clarifications into the SCSI-2 document during the final editing. 12. Meeting Schedule The next meeting of X3T9.2 will be August 20-21, 1990 at the Clarion Harvest House in Boulder, CO (303) 443-3850 hosted by Hughes Aircraft Company. Please mention the X3T9/Hughes meetings when making reservations to receive the group room rate of $85.00 single or $95.00 double, plus tax. The cut- off date for reservations is July 19, 1990. The plenary meeting schedule is: Date Location Host -------------------- ------------------------ --------------------- August 20-21, 1990 Boulder, CO Hughes (*) October 15-16, 1990 Ft. Lauderdale, FL Boeing (*) December 3-4, 1990 San Jose, CA AMD (*) February 18-19, 1991 Dallas, TX EDS April 22-23, 1991 St. Petersburg, FL AMP (*) These meeting locations have changed and were selected at the February 23, 1990 X3T9 meeting. These meetings are being arranged by the individual hosts without Classic Conferences. Please do not call Classic Conferences regarding any meeting after the June 1990 meeting. 12.1 General Working Group Schedule for 1990 The Working Group Schedule is as follows: Date Location Host Comments ------------------- -------------- --------------- ---------------- Jul 9-13, 1990 Rochester, MN IBM Sep 4-7, 1990 Denver, CO Storagetek {no CAM meeting} Oct 29-Nov 2, 1990 Austin, TX TI These week-long meetings are to be divided as follows: Monday SCSI CAM Committee Monday Cable Specification Working Group (July 2:00 pm) (No Monday meetings in September -- Labor Day.) Tuesday-Wednesday SCSI Working Group Tuesday-Wednesday HIPPI Working Group Thursday-Friday Fiber Channel Thursday-Friday Fiber Channel FC-0 Thursday-Friday IPI-2 Working Group 12.2 Financing Plenary Meetings John Lohmeyer said that he had received a preliminary draft of new rules to permit Technical Committees to collect funds for specific purposes. He said that he is pleased with the document and it should empower us to finance meetings through meeting fees or contract with an editor to do editing work. It precludes buying capital items, such as computers or printers. We could rent such equipment, however. 13. Adjournment The meeting was adjourned at 9:52 a.m. on Tuesday, June 19, 1990. Meeting Attendees - Enclosure (1) Name Status Organization ------------------------------ ------ ------------------------------ Mr. Wayne E. Roen P 3M Company Mr. Robert C. Herron A 3M Company Mr. John Knudson S 3M Company Mr. James Fiala S 3M Company Mr. Al Wilhelm P Adaptec, Inc. Mr. James R. Schmidt V Adaptec, Inc. Mr. Jim Kubinec O Advanced Micro Devices Mr. Gary W. Arakaki O Advanced Micro Devices Mr. Charles Brill P AMP, Inc. Mr. C. Edward Reynolds A AMP, Inc. Mr. Bob Whiteman A AMP, Inc. Mr. William R. Sopchak A Amphenol Interconnect Mr. Jan V. Dedek P Ancot Corp. Mr. Robert Otis O Apple Computer Mr. Peter M. Blackford P Astro Cable Company Mr. Joe Lawlor P AT&T Mr. John Ellis P Berk-Tek, Inc. Mr. John F. Osborn A Berk-Tek, Inc. Mr. Joseph P. McCart V Cables Plus Mr. John Contreras O California Peripherals Mr. John Guennewig P Cinch Connector Mr. Bharat Shah A Cipher Data Products, Inc. Mr. Thomas Wicklund P Ciprico Inc. Mr. Fred Hengelhaupt O Connective Technologies, Inc. Mr. Stephen R. Cornaby P Conner Peripherals Mr. Wayne Sanderson P Control Data Corp. Mr. Sam Pendleton P Data General Corp. Mr. Mark P. Pearson P Datacopy Corp. Mr. Rudy Stalzer A Digital Equipment Corp. Mr. Del Shoemaker L Digital Equipment Corp. Mr. Ed Haske A Distributed Logic Corp. Mr. Stephen Fitzgerald P Eastman Kodak Co. Mr. Paul Hanmann P Emulex Corp. Mr. Chuck Micalizzi A Emulex Corp. Mr. I. Dal Allan P ENDL Mr. Robert Liu P Fujitsu America, Inc. Mr. John C. Onia A Fujitsu America, Inc. Mr. Terry Maezawa P Furukawa Electric Amer, Inc. Mr. Kenneth Post P Future Domain Mr. Kurt Chan P Hewlett Packard Co. Mr. Mike Peper A Hewlett Packard Co. Mr. Oscar Kornblum A Hirose Electric U.S.A. Mr. David McFadden P Honda Connector Mr. Tom Kulesza O Honda Connector Mr. George Penokie P IBM Corp. Mr. Steve Kappes O Interphase Corp. Mr. Paul Jackson A Iomega Corp. Mr. Ted Dubbs O ITT Cannon Mr. D. Shaff A JAE Mr. Robert Bellino P Madison Cable Corp. Ms. Donna Pope V Maxoptix/Maxtor Corp. Meeting Attendees - Enclosure (1) Name Status Organization ------------------------------ ------ ------------------------------ Mr. Lawrence J. Lamers P Maxtor Corp. Mr. Bob Masterson P Methode Electronics, Inc. Mr. John Spongr A Mitsubishi Electronics Amer Mr. Jim McGrath A Molex Inc. Ms. Julie Duvall V Molex Inc. Mr. Gary Murdock S National Semiconductor Mr. John Lohmeyer P NCR Corp. Mr. Manfred Hoyer V NCR WPD Augsburg Mr. Thomas R. Marks O NEK/Helix Cable, Inc. Mr. David Filpus P Northern Telecom, Inc. Mr. Dennis P. Trupski P Olivetti Mr. Ed Cieniawa A Panduit Corp. Mr. John Emerson O Prime Computer Mr. Gerald Houlder A Seagate Technology Mr. Gene Milligan A Seagate Technology Mr. James V. Reitman V Sine Companies Mr. Michael Britch P Standard Microsystems Corp. Mr. Vit Novak O Sun Microsystems, Inc. Mr. Tom Gardner O Technology Forums, Ltd. Mr. D. W. Spence P Texas Instruments Mr. Edward R. Schurig O Texas Instruments Mr. Harvey Waltersdorf P Thomas & Betts Mr. James Patton P Visqus Mr. Erik Jessen O Western Digital Mr. Greg Leonhardt A Zenith Data Systems 76 People Present Status Key: P - Principal Member A - Alternate Member L - Liaison O - Observer S,V - Visitor