[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: cp command
From: |
Patrick J Clas |
Subject: |
Re: cp command |
Date: |
Tue, 19 Jul 2005 11:04:50 -0400 |
Thanks for your quick reply. Specifically, I am attempting to preserve
the timestamps of files with the -p flag. The script I'm running copies
many files and some of them have permissions that prevent the preserve
from working. The problem is that cp returns 1 for this failure and I
fail even though it's an acceptable condition. I guess I'd like to see cp
return a different return code so I didn't have to parse the string output
of the command, or perhaps an additional flag that says we don't care if
the preserve fails. I do care about severe errors like the copy failing
completely, but if the preserve fails, the copy still could've succeeded.
Patrick J. Clas - Software Engineer - zSeries
IBM, Endicott
Phone: (607) 429-4425 T/L 620-4425
Office: 256-3 X007
address@hidden (Bob Proulx)
07/19/2005 10:41 AM
Please respond to
bug-coreutils
To
Patrick J Clas/Endicott/address@hidden
cc
address@hidden
Subject
Re: cp command
Patrick J Clas wrote:
> Is it intentional that no matter what the failure is, cp always returns
1
> instead of something meaningful for various failures?
You say that as if returning a 1 upon a failure is not meaningful.
Therefore you must have something in particular in mind. Could you
expand a little on what you are thinking here? Is there anything in
particular that is inadequate?
In general there are many possible ways for a program to fail.
Personally I believe that trying to enumerate all possible failures is
not a good way to do things because it is never good to try to
enumerate an unbounded set. And also on different operating systems
there will be different failure modes available. This makes the
handling of all possible errors impossible to keep the same across all
platforms. Better to keep the model straight forward.
Bob
Re: cp command, Philip Rowlands, 2005/07/19