Hi Jack,

 JS> I am having a difficult time believing that FDISK recognized you had a 
 JS> type B partition, yet refuses to delete the partition.  

I have followed this thread in the background until now, however a past
experience of mine MANY YEARS AGO was very similar and may provide a clue. In
my case it was with SCSI Host Adapters and drives, but there are strong
similarities to the problem here. In all cases I was using OS/2 FDISK, because 
my DOS FDISK was V3.3 (I told you it was a long time ago......;-)) and it
would not touch HPFS partitions...

After moving a drive from one SCSI HA (an Adaptec 1540) to another (a Future
Domain), it ran fine as it was set up, but eventually I needed to re-partition 
the drive, and this is when my hassles started. I could run OS/2 FDISK and
remove the partitions, even save the changes, but upon system restart ALL
ORIGINAL partitions were still there exactly as they had been defined. It was
as though the partition table was write protected, the rest of the disk could
be written no problem...

Eventually I tried another HD and that had the EXACT same problem, so I then
went back to using the Adaptec HA and that worked fine for partition table
changes. Back on the FD HA I could still not WRITE a partition table. 

After much playing around I discoverd that the issue related to a strapping on 
the Adaptec that enabled LARGE drive support for booting under DOS. From
memory, I think this changed the way the SCSI HA mapped the drive from 16
sectors, 16 heads/cyl to 64 sectors, 32 heads/cyl. By setting it back to the
OLD value (16/16) the partition table could be written on BOTH HA's... 

The sequence of events was that I had changed the Adaptec HA from the OLD head 
mapping scheme to the NEW mapping scheme without re-writing the partition
table, but this apparently did not worry the Adaptec, it was able to handle
the situation. The down side is that it caused real problems for the FD when
WRITING to the partition table of the drive, but it could still read it! Of
course with OS/2, once the drivers were loaded the BIOS was never used again,
so the partition data areas were able to be addressed fine for reading... Once 
I cleared all partition info using the OLD mapping scheme, changed the Adaptec 
back to the NEW scheme, wrote a NEW partition table, then BOTH HA's were happy 
with writing the partition table again...

This may be a rather long shot, but I wonder if it is possible that the drive
partition was created with LBA turned off (or on), but subsequently LBA has
been turned on (or off) and the partitions changed again. How about reversing
the LBA state and trying to partition it again...

I hope some of this helps.........pk.


--- Maximus/2 3.01