On 19 Jul 2005 10:43:26 -0700, in bit.listserv.ibm-main you wrote: >Actually the unused part of a disk address is the BIN number, also from >the noodle snatcher. M is used for the extent number now. > >A full disk address is: MBBCCHHR > >Today, the BB field has to be zeros. Using those two bytes would >effectively allow a 4G cylinder device, however there are some caveats >to that. > >A disk record address is CCHHRKDD. Using the BB field would allow there >to be multiple disk records with the same CCHHR on a device. They could >get around that by having the restriction that a DEFINE EXTENT command >can not cross bins. Actually, there is no BB field in the DEFINE EXTENT >command. They could use the reserved byte 6 and unused byte 7 to hold >the BB data. The LOCATE RECORD command would then be local to the >defined bin. > >That would allow for a 1 tera-track device, or about a 56 peta-byte >device if we still limit the bytes per track to 56,664. > >All of this is just speculation; we will just have to wait for IBM to >decide.
Why not just decide what logical file formats (sequential, byte stream, keyed, direct, etc.) are needed, come up with the control blocks to handle them which also would be designed to handle current files and start the move to FBA? The control blocks would be designed to handle longer names, Unicode and megabyte or greater record lengths. Existing code would slowly (5 - 15 years) die. We now have consumer PCs with 200 plus gigabytes so the mainframe is lagging in this area. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
