[<<]Message[>>]    [<<]Author[>>]    [<<]Subject[>>]    [<<]Thread[>>]

Number   :  4079
Date     :  2003-03-19
Author   :  Garry Deane
Subject  :  Re: using xxcopy together with zip/arj compression
Size(KB) :  2


--- In xxcopy@yahoogroups.com, anders thoresson  wrote:
> Using xxcopy for backup puroposes, one of the things I
> want to do is to backup my most crucial files to a 250MB
> zip disc. To extend the capacity of the zip-disc, is it
> possible to set up a batch file that lets xxcopy work 
> closely with pkzip/arj?
> 
> Written in pseudo-pseudo-code, I think something like
> this would work, but is there a better way to do it:
> 
> First, backup all files, make zip-file out of it and
> put on the zip-disc. Then on every shutdown:
> 
> 1. Backup changed or new files to a directory on my
> hard drive.
> 2. If any files exists in the directory or it's
> sub-directories, update zip-file on zip-drive.
> 3. Delete backuped files and directories from
> temp directory.
> 4. Shut down.

That pseudo code approach looks fine. What you need to work out is 
how you will select files for putting into the temp backup directory 
because there are no files there for xxcopy to compare against. The 
usual and easiest approach is to use the archive bit to select the 
new or changed files. Generally speaking, the archive bit gets set 
whenever you create or modify a file but it is not bullet-proof 
because other programs may also manipulate the archive bit. Use the 
xxcopy switch /M to select files if the archive bit is set then clear 
it after the copy.

However if you're going to use this approach, you may not need to use 
xxcopy. Your zip program probably also has a switch which will only 
add files if the archive bit is set and clear it after the zip 
operation. If your zip program allows for both updating (don't add 
files if same file is already in the zip) and using the archive bit, 
then you can just use that without the intermediate step of using 
xxcopy.

A more complex approach which is more bullet-proof for this sort of 
operation has been posted by Kan some time ago. It involves creating 
a catalogue of all the files on the computer using /SZ0 (now /TR0) 
which creates a directory structure and file list of zero length 
files. There's an example of this in technical bulletin #11. This 
reference directory is then used to compare the timestamps of updated 
files and set or clear the archive bit of the source files depending 
on whether the source files are different to the reference directory 
files. Those files with the archive bit set can then be copied to the 
temp backup directory. Be warned however that this is a reasonably 
complex set of operations so is not for the faint hearted.

One last point is that I would keep a copy of the zip file on the HDD 
and do the updates to that file. Once it has been updated, simply 
copy it over the previous version on the zip drive. This will save 
time accessing the relatively slow zip drive.

Garry



This message if part of XXCOPY's message Archive.
The archive contains all the messages posted at Yahoo!Groups: XXCOPY.