JEAN PARROT wrote in a message to LEONARD ERICKSON:

 JP>   Good morning to you Leonard, how are you today ? 

 JP> The two HDs are on the same IDE cable, M & S and the two CDs
 JP> are on the other IDE cable, again M & S.
 JP>  
 JP> Is the D:\ HD at ATA 33 slowing things down at all ?

 LE> I'd suspect so. Also, you'll get better performance if you have each
 LE> HD as Master with a CD as slave on the same cable. That'll maximize HD
 LE> to HD transfer speeds. 

 JP> Are you not going against the grain of the wider opinions
 JP> here ? I got back into the box here to change all this to
 JP> the HDs on one IDE and the CDs on tother because all were
 JP> suggesting that it was the way to have it when before I had
 JP> it the way you now suggest.   Strange !

It depends on how you plan to use things.  The key point is to consider
that an IDE channel can only do one thing at a time.  So,  if you try to
perform an operation that involves two devices on the same cable,  one
device is going to have to finish before the other one can go into action. 
So putting two HDs on two different cables _does_ optimize HD-to-HD
operations.  This is why SCSI is used in higher-end server applications, 
you can issue a command to a SCSI device (say,  to seek to a certain track
on a HD) and the device then releases the bus to allow for other things to
happen.  You can have a bunch of overlapping operations and overall better
performance that way.

---