[pups] RL and RK Last Cylinder Warning (RE: VTServer)

[Jay Jaeger]

Something I thought is worth warning folks about.  For RL01, RL02, RK06 and 
RK07 (at least), the last cylinder is RESERVED for bad blocks and the pack 
serial number.

Except for maintaining the bad block list, software should not write there, 
and should at least preserve the serial number (and its mirror copies that 
are usually present) when doing so.

The current implementation of VTServer (in particular, the code in the "hk" 
driver in pdpvtstand) does not appear to take that into account.

The problem is that if you take a *full* pack image from a pack and then 
restore that *full* pack image to a different pack, you will wipe out the 
latter's bad block info (and perhaps write in some areas that are known not 
to be very good).

Traditionally, Unix variants dealt with this by insisting that you use 
error free (Suffix -EF in the DEC part number on the pack) disk 
packs.  However, if you are using real hardware you may no longer have 
those available.

Just a "heads up".

(PS:  FYI I am currently working on a flavor of the "vt" driver that will 
work over a DR11C.  With the proper cabling (it takes different cables for 
input and output) you can manage over 1MB per minute that way -- about 20 
times faster than a serial port.  The code is pretty simple and should work 
on any old "plain jane" parallel port.  (I do not know how it would behave 
on a modern serial port)).

I had this working a long time ago, and VTServer prompted me to dredge it 
up again and test it.  It still works with my PIII 350 under Win98 (it is 
old DOS code that accesses the old style parallel port directly).  Within a 
few weeks I expect to provide a "pt" device driver for pdpvtstand, and 
cabling information and some source code for the PC side (which would be an 
example only).  The protocol as I currently use it has no 
checksum.  However I plan to modify "copy" to provide that information so 
that the results can be checked after the transfer.

Jay Jaeger
---	
Jay R. Jaeger					The Computer Collection
cube1 at home.com			visit http://members.home.net/thecomputercollection


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id NAA27320
	for pups-liszt; Wed, 4 Apr 2001 13:11:21 +1000 (EST)
	(envelope-from owner-pups at minnie.cs.adfa.edu.au)