Hi Richard,

I was a customer at the time ADSM came out, so I was not privy to the
original design. However, it is likely that the answer lies, at least in
part, with RESTORE syntax, where we need to know whether the file spec is
a directory or file (i.e. do we search high-level names or low-level names
in the TSM database). Also, it is possible to have a file named 'A', back
it up, delete it, create a directory called 'A', then back it up. So now
you would have a directory named 'A' and a file named 'A'. The terminating
backslash helps resolve the ambiguity.

Another case is when you wish to restore to an alternate location. You
need to specify the backslash on the target file spec (if it does not
exist on the client file system) to indicate that the target is a
directory, not a file name. For example, if I have a directory named DIR
with nothing else in it except for a file named FILE, then I could say,

   res c:\dir\file c:\dir\x

or

   res c:\dir\file c:\dir\x\

In the first case, the file would be restored as name 'x' in the 'DIR'
directory.

In the second case, the file would be restored as name 'file' in the DIR\x
directory.

Requiring the backslash on the backup file spec when dealing with a
directory helps keep the syntax consistent.

Regards,

Andy

Andy Raibeck
IBM Tivoli Systems
Tivoli Storage Manager Client Development
e-mail: araibeck AT us.ibm DOT com

The only dumb question is the one that goes unasked.
The command line is your friend.
"Good enough" is the enemy of excellence.





Richard Sims <rbs AT BU DOT EDU>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
08/14/2001 04:40
Please respond to "ADSM: Dist Stor Manager"


        To:     ADSM-L AT VM.MARIST DOT EDU
        cc:
        Subject:        Re: Client Backup Issue



>In the schedule settings, you should change the OBJECTS from:
>
>   C:\TSM Backup
>
>to
>
>   C:\TSM Backup\
>
>unless you only want to back up the single file or directory named "TSM
>Backup", but I doubt that is what you want.

Andy - I have to ask: why does the software require customers to add
       a slash or backslash (depending upon architecture type) to the
end of a specification for the client to know that the object is a
directory?  As a programmer, I would write the software to simply test
for the actual object being a directory or not, rather than silently
fail to process the directory.  That is, if the customer said to back
up XYZ, I would do so upon discovering the existence of the object,
be it a file or directory.

The slash-backslash requirement is counter-intuitive for customers in
general, where operating system commands never require adding a special
character to the end of a spec to say that the object is a directory:
such commands simply proceed to honor the invoker's request and process
the object as found.  This special TSM requirement needlessly confuses
many customers, and is a pet peeve with the product.

  Richard Sims, BU