is cable select not reliable?
this is the 1st time I used that setup
Well, AFAIK cable select just sets the drive next to the controller as slave,
the one at the end of the cable as master. Also AFAIK, both drives must be
cable select. So it should work. But I've never used it. Your problem may be
caused by cable select not working right, that's why I suggested going back
to master / slave.
There is some confusion about where your boot drive should be -- I've read
that BIOS will try boot first whatever it's told to boot first, say a floppy
drive, and if none of the specified devices boot, it will go to the first
bootable partition on the master attached to the primary controller. This
latter habit makes boot managers possible, BTW.
If there is more than one HDD, BIOS should give you the option of setting HD0
or HD1 as the boot drive (in the boot sequence setting.). Normally, the
master on the primary controller will be seen as HD0, and it should be the
boot drive. I'd arrange things so that is the case. I don't like departing
from common practice. IOW, I won't use a new feature (such as cable select)
unless I was sure it would work at least as well as what I was used to. If
the machine came supplied with two HDDs, both set to cable select, that would
be the proof it worked as it should.
BTW, at least once I found that if the master is not at the end of the cable,
you can have problems. This was on an older machine with a BIOS dated 1997,
so things may be different with your BIOS. Anyhow, I'd make the boot drive
the master, and put it on the end of the cable.
Some people have reported that mixing HDDs and CD drives on the same cable
causes problems. "It's all rather confusing really." (Neddy Seagoon.)
<short rant:>
The variations in the problems reported leads me to believe that the
underlying cause is a lack of industry-wide and enforced standards in BIOS
design. Of course, if we had true smart drives, whose read and write
operations were totally independent of the operating system, none of this
would be a problem. BIOS would simply query the attached drives for bootable
partitions, list them, and either default to some designated OS, or boot with
your choice. Which would make both BIOS and OS a hell of a lot simpler and
more reliable IMO. There was the beginning of such a technology, pioneered by
Commodore and Amiga, among others, but along came the unholy alliance of IBM
and Microsoft, and we got the PC -- an inherently crappy technology, which in
essence is a kludge of patches. It's amazing it works at all, really.
<end short rant>
HTH&GL