This message is from the T13 list server.
On Thu, 9 Jan 2003 11:24:23 -0800, Gana Pat wrote: >Especially when it comes to ATA external storage boxes and ATA drives on >RAID controllers, there is no way that the drive vendor's FW update feature >can be used without connecting the drives to legacy controller. If not >today, may be some day in the near future, when the drive FW becomes >complex and updates are mandated, we need a standard mechanism for FW >update. Hmmm... Now I'm real curious... "near future, when the drive FW becomes complex"... Lets see today's drive has a multitasking OS, some have multiprocessors, most have 6 to 20 tasks running all the time, many have hardware state machines also running all the time... and this isn't "complex" today? Where is this additional complexity going to come from? If that additional complexity happens then I can only assume that the anti-SCSI folks will have been sucessful in creating some kind of SATA monster that replaces SCSI but really has all the SCSI overhead and features the anti-SCSI people don't like today, plus some things we don't know about yet or things the SCSI people have rejected as really bad ideas. Some day someone will realize that a disk drive is nothing more than a "memory chip" and that the disk drive doesn't need all this additional crap. A disk drive really needs only four commands: identify yourself, read data, write data and maybe go to low power mode. That's it - nothing more is needed. All that additional crap we find in today's disk drives does nothing but make the drives less reliable. I often wonder what would happen if it was possible to build a "disk drive" that had the same form factor, same power requirements, same performance and same price as today's drives but was entirely solid state (no moving parts) and with reliability 100 to 1000 times that of today's drives? Would we still have people proposing new SMART subcommands? What would happen if these new drive had no firmware to update? What if power management was entirely automatic? Of course the anser to my questions is: Yes, there would be new SMART subcommands, new power management commands, firmware would have to exist (because drives had always had firmware!) - all this is required because it is never possible make anything less complex in our industry. Complexity is good, complexity is required, complexity is cheap, ..., forget about making simple products that are reliable. Hale *** Hale Landis *** www.ata-atapi.com ***
