What does Saving As do?

["Herbert M. Harrell" <hm...@hassoc.com>]

Greetings all!

Does anyone know exactly what at a minimum changes in a MS Office file 
when it is "Saved As"?

Think of a situation where you download a Word document, which is 
automatically opened in Word. Without making any changes, you save it 
under its original name to your desktop.

Your copy is no longer a byte-for-byte duplicate of the original. 
Certain Summary Information properties have changed (e.g. edit time, 
last author, revision number).

Are the differences between the original and your copy confined to 
Summary Information?

That is my precise technical question. An answer would be much 
appreciated.

I'm hoping the answer is "yes", but in case it is "no", I'd appreciate 
some indication of where else differences could occur.





In case anyone is interested, here is the background:

If the differences between original and copy are confined to Summary 
Information, then RAINER KLUTE's WriteAuthorAndTitle.java could be 
adapted to test for functional identity of your copy and the original. 
(See the examples section of the POI source tree.)

Here's how it would work:

You would upload your copy back to server where it came from. The 
server would create near-duplicates of the original and what it gets 
back from you, but set in both instances the SummaryInformation 
properties to null; these near-duplicates would then be compared byte 
for byte.

Why would anyone want to do this? Suppose you were a subcontractor on a 
multibillion dollar building project. If the main contractor wanted to 
position himself to be able to prove in court, should the need someday 
arise, that you consciously received a true and undamaged copy of a 
specific version of an engineering specification on a today, he might 
ask you to go through the upload exercise. Of course, he would make it 
as convenient for himself and for you as possible, but I'll leave out 
that part ...


Herb Harrell
Harrell Associates Incorporated
601 King Street
Alexandria, Virginia 22314

+ 1 703 518 5550


---------------------------------------------------------------------
To unsubscribe, e-mail: poi-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: poi-user-help@jakarta.apache.org

RE: What does Saving As do?

[Michael Zalewski <za...@optonline.net>]

... And, in the case of Word, especially if the document is Quick Saved, the
stream will change.

Another thing that will change in Word and in Excel is if you simply print
the document. Both Word and Excel will change the document stream, depending
on the resolution and the driver characteristics of the printer you select.

Even the structure of a word document can change. In this case, when you
save the Word document after downloading, even the page breaks can change,
because of slight variations in the actual font size used by the printer
driver for the default printer you have selected. This is particularly true
if your default printer has a resolution of 600 dpi or 1200 dpi, but the
document you upload was created with a printer of lower resolution. It turns
out that 10 pt Arial or Times Roman renders as slightly larger on a 600 dpi
printer vs 300 dpi. So if you save a Word document while a different
resolution of printer is your default printer, you may find that soft page
breaks are either inserted or removed. A printer of higher resolution tends
to insert page breaks, or move them to previous lines.

Also, I don't get how uploading the saved file actually proves anything,
especially if the subcontractor disagrees with the main contractor. If the
subcontractor simply uploads the file, the main contractor can say you must
have seen this version of the document - and produce the copy that the
subcontractor uploaded. The bytes making up the file that the contractor
uploaded might be completely different. But that would be OK as long as the
main contractor saved the copy that the subcontractor uploaded. Of course,
the GC still has to be able to show that the copy came from the
subcontractor, not from some other source. And that no subsequent changes to
the specification had been agreed upon.

Would it not be better to use a secure transfer? Maybe coupled with some
algorithm such as Kerberos?

-----Original Message-----
From: news [mailto:news@sea.gmane.org]On Behalf Of Chris Nokleberg
Sent: Monday, August 16, 2004 3:14 PM
To: poi-user@jakarta.apache.org
Subject: Re: What does Saving As do?

Herbert M. Harrell wrote:
> Think of a situation where you download a Word document, which is
> automatically opened in Word. Without making any changes, you save it
> under its original name to your desktop.
>
> Your copy is no longer a byte-for-byte duplicate of the original.

In fact you do not need to even save the document for there to be changes.
For example, simply *opening* a PowerPoint document will modify the user
name stored in the CurrentUserAtom record, unless you have the file marked
as read-only.

Chris



---------------------------------------------------------------------
To unsubscribe, e-mail: poi-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: poi-user-help@jakarta.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: poi-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: poi-user-help@jakarta.apache.org

Re: What does Saving As do?

[Chris Nokleberg <ch...@sixlegs.com>]

Herbert M. Harrell wrote:
> Think of a situation where you download a Word document, which is
> automatically opened in Word. Without making any changes, you save it
> under its original name to your desktop.
> 
> Your copy is no longer a byte-for-byte duplicate of the original.

In fact you do not need to even save the document for there to be changes.
For example, simply *opening* a PowerPoint document will modify the user
name stored in the CurrentUserAtom record, unless you have the file marked
as read-only.

Chris



---------------------------------------------------------------------
To unsubscribe, e-mail: poi-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: poi-user-help@jakarta.apache.org