June 12, 1989 The CAM (Common Access Method) Committee meetings of May 8-9, were hosted by Ballard Synergy in Cupertino. The following attendees were present: M. Aarons EMULEX D. O'Shea FUTURE DOMAIN D. Allan ENDL R. Perry RHETOREX J. Armstrong NCR J. Polonsky SMS C. Ballard BALLARD SYNERGY T. Ram NETFRAME SYSTEMS S. Betts NCR T. Reichert INTERACTIVE SYSTEMS P. Boulay LMS OSD M. Robinson SEAGATE F. Burgess SEAGATE L. Robinson STORAGE DIMENSIONS D. Bursky ELECTRONIC DESIGN T. Rossbottom TEXAS INSTRUMENTS J. Chen BUSTEK L. Sass CALIPER J. Chen CIRRUS LOGIC S. Shah WESTERN DIGITAL B. Duran CALIPER T. Shea ADAPTEC D. Flower COMPAQ D. Skinner NCR J. Hamada MITSUBISHI ELECT'S S. Smyers AMD R. Hoehnle SEAGATE B. Snively SUN MICROSYSTEMS W. Ihde LMS OSD J. Spongr MITSUBISHI ELECT'S M. Jerbic HEWLETT PACKARD T. Stockwell MICROSOFT J. Katzung APPLE R. Tewell COLUMBIA DATA J. Lawlor AT&T BELL LABS T. Treadway DPT C. Linsley SMS D. Trupski OLIVETTI P. Manley SYQUEST TECHNOLOGY A. Turenne ARCHIVE G. Marazas IBM M. Van Scherrenburg NATIONAL SEMI G. Milligan CDC IMPRIMIS J. Webb SYTRON R. Monsour STAC C. Yang EMULEX T. Murray NCR E. Young ARCHIVE S. Neuner HITACHI COMP PRODS Working Group Reports The efforts by working groups are summarized here. Participants interested in the specific of the working group activities should call the leader of each effort. ATA/EATA Tom Hanan's comments on ATA were reviewed, and he is going to provide new sequence and timing information that will help comprehension. Tom is also compiling a list of all commands used by vendors at present so that we can assign opcodes to Vendor Unique or Reserved categories. If your company has a 40-pin ATA, send Tom (Fax: 714-553-8152) a list of your commands and their opcodes. If you do not, you just might wind up with your opcodes defined as Reserved, and not supposed to be used. A general concern over coupling ATA and EATA in the same document means that following the rev in this mailing, there will be a change in documentation to make it clearer that the two are separate, and that EATA is optional and an alternative. The ATA is to be specific to the 40-pin implementation. Bob Monsour reported that the Quarter Inch Cartridge committee is looking closely at EATA as a means of attached tape drives. OS/2 Tracy Murray had a few problems in getting the meeting spot arranged, but the meeting itself was successful in that there were three representatives from Microsoft, and the ball is rolling on the list of issues which need to be addressed for OS/2. Microsoft has assigned a representative to work with the CAM Committee. Unix Todd Reichert held a small gathering at which a Sun Microsystems document was discussed as the basis for a Unix OSD. Until this document has been made more generic it will not be distributed to the Committee, but remain within the working group. Meeting Jerry Marazas made a short presentation to clarify his comments at the last meeting re the OS/2 Interrupt Manager. The Interrupt Manager has not really changed, it is more the philosophy. From one device type per Host Bus Adapter per driver to n device types per Host Bus Adapter per n drivers. The changes made are in the BIOS rather than the Interrupt Manager. Jerry Armstrong, Ron Perry, Shishir Shah, and Rick Tewell are merging the SIM, DOS OSD and Transport functions. Ron reported the changes made to the DOS OSD in response to earlier committee criticism. The fact that the OSD has to know about all the devices in advance (via CONFIG.SYS statements) was not well accepted. There was a general feeling that a user should not be forced to do this in all environments. The concept of whether there should be a static configuration at all was debated, and ended with general concurrence that it is desirable, but there was no agreement on the technique to provide this capability. It is likely vary by platform and OSD. Tom Shea pointed out that OS/2 on the AT makes no provision for a static configuration. Rick Tewell is to provide a precis of the problem and why a static config capability is needed to Microsoft. In Rick's presentation he explained that the degree of overlap between the OSD and the routing services of the Transport module make it desirable that the two be merged into an OS/Transport service. Providing the generic, aka short block, requests in the OST bothered Tom Shea as it implied there would only be a single entry point. If a vendor chose only to provide the routing functions and no short block services, a mechanism would have to be developed and supported by all HBA manufacturers to map across to short block software provided by other companies. Tom preferred multiple entry points so they could be independent. The general decision reached was to treat short block services as a device driver. A device vendor could provide a driver which supported the short blocks, and all would share a common router. It was agreed that there would be only one router in a system (if there is more than one, each would be oblivious to the other). This is to ensure that there is only one path to a device, and requests get properly serialized. John Webb is to supply a list of recommended short block services for tape and Patrick Manley is to do the same for disk. Jerry Katzung presented an overview of the improved services offered under Release 7 for the Macintosh. One of his biggest problems comes from having to support the old mode as well as the new. This drew laughter from all of those who have been suffering the same kind of pain in the IBM PC world. BBS Activities Rick Tewell is collating ideas to get on paper for the OSD/Transport. Anyone with direct input can reach him via the Columbia Data BBS (407-862-4724). Contact Ed Rauch at 407-869-6700 to get a user ID. Todd Reichert reminded everyone that if they want to participate in the Unix Working Group activities, to upload their E-Mail number to the SCSI BBS so they can be added to the E-Mail distribution list. Area 13 of the SCSI BBS (316-636-8700) is the clearing house between members working on proposals. Proposals etc will be in the Files section and short exchanges are in the Message section. If you think something should be on the board but is not, you may want to check Area 9 also. Complaints were heard at the meeting re the difficulty of getting into the BBS - either no ringing or connecting at 1200 instead of 2400 baud. The latter appears to be a modem problem. On occasion, I have found it helpful to go back to AT&T for dialing the SCSI BBS. Access in my area is by a prefix of 10288. Future Meetings: Air fares are the most expensive part of attending meetings, and in recent months we have all noticed a general increase in costs as the airlines cut back on their discounted travel programs. The airlines which have agreed to provide special air fares for all CAM meetings (as well as other ANSI meetings and working groups) are United and Northwest. The arrangements for each airline are the same. - Call a special 800 number which connects you to the Meetings desk. - The Meetings desk has limited hours, so no late night calling! - Supply the Travel ID Number for the airline you are calling. - Receive a flat 40% discount on regular air fares. - Receive an additional 5% on discounted fares. - Make arrangements at least 7 days in advance. - You can be ticketed by your own agent. Travel agents can use the number but they do not seem to like calling the Meetings desk - it takes longer than making reservations on their terminals and the airfare is lower (smaller commission). Airline Phone Number Meeting ------- ----- ------ ------- United 800-521-4041 413TA All Northwest 800-328-1111 15201 All Our hosts are committed to a minimum number of room nights which are used to offset the cost of the conference rooms. Please make sure the hotel is given the name of the host organizing the meeting, even if it means placing the call yourself. Date Host City Hotel Hotel Phone ---- ---- ---- ----- ----------- 7/12/89 AT&T Chicago Ramada O'Hare 800-272-6232 * 8/9-10 Emulex Costa Mesa Red Lion 714-540-7000 9/ 8/89 Imprimis Oklahoma City tba 10/ 6/89 DPT Orlando tba 11/ 1/89 Future Domain Irvine tba * Note that this is a 2-day meeting on the assumption that we will have to review documents and take voting decisions on level of completion. I. Dal Allan