> If an I/O error occurs upon writing to a file (as opposed to opening it),
> then cp will remove the file.  This is true even if the file is
> a special device or named pipe.

I believe that this behavior is a bug in cp.  In fact, I have a
message in my inbox which notes that traditional cp does not remove
the file on write failure.  Your message reminds me that I should
confirm that POSIX.2 does not require this "feature" (I don't think it
does), and arrange to get cp patched.

	--jtc