DM> disk mangler is different. It's a software
DM> rather than a hardware/firmware implementation.

BL> I shudder at the idea of relying on software
BL> to access the disc with the software on it.

DM> In it's time, it worked remarkably well. I had installed
DM> it ion several PC's, and none ever caused me any grief.

The main irritation with it compared with LBA is where you want to
say move to OS/2 from a DOS/Win config etc. LBA is rather cleaner.
The main downside with going that route is that it can fang you if
you have a >528MB drive, have upgraded your motherboard to one with
LBA support, and then you want to step back to your old motherboard
without LBA support for some reason like the new MB has some warts etc.

BL> But my BIOS knows about LBA, so where does that leave me?
BL> Do I need a special VLB board or just an ordinary one.

DM> Ordinary IDE card should do fine.

BL> I need the one for four drives so I can run the CD
BL> separately. I thought I needed a VLB card for that.

DM> Not necessarily, although the bulk of
DM> the 4-HDD controllers seem to be VLB based.

No real need to have it all on one card, any even pure IDE paddle card
which can be jumpered to the secondary etc IDE port stuff is fine.

DM> I'm basically in the same boat as you, viz the CD ROM drive. Currently
DM> I've been using a yum-cha audio card, with the audio bits disabled, to
DM> access the CD ROM. But it seems to conflict occasionally with OS/2 and
DM> perhaps the MediaVision PAS card, leaving the system unstable.

Why not try a simple ISA IDE paddle card which can be
jumpered to the secondary IDE port address etc. Most can be.

BL> Can't I enable LBA for the big drive but access
BL> the small one non-LBA? I back up to the small drive.

DM> Yes, you can do that - LBA is implemented on a per-drive basis

DM> Some of the newer BIOS's also sport a setting called LARGe, in addition
DM> to LBA and standard modes. I have no idea if this is an alternative
DM> method of accessing the entire HDD space. Maybe Rod would know.

LARGE is a different way of getting around the 1024/16 limit, allowing
more than 16 heads, but the BIOS translates that into a cylinder/head
combination of >1024 cylinders and <= 16 heads, usually by doubling the
heads and halving the cylinders. So DOS and the bios operates in terms
of CHS values of 16 heads and that is translated
by the bios into >1024 cylinders and <=16 heads before the command
goes out on the cable in CHS values to the drive itself.

This one is only relevant for OSs which use the BIOS for the
drive. Modern stuff like OS/2 which replaces the BIOS with its
own code once its properly booted has to be able to do that CHS
translation itself, and LBA is a rather cleaner route with those.
@EOT:

---