Sorry for the late reply. Been on a plane moving about.

You might be correct now. Just know that way back when in the DOS (for you
newbies that is before Win9x <G>) days when I used Norton Speed Disk, they
(the Norton group, when Mr.(Mr. out of respect)  Norton still own/ran it and
I trusted software that had his name on it.)  were very specific to back up
the hard drive BEFORE during a defrag because  of the possibility of lost
data during defragging for reasons I have described previously. Maybe things
have changed but I personally have not read that to be true. Does not mean
it is not accurate now, just have not seen documentation to contradict that
last documentation I have read from Peter Norton (by the way the Win98
Defrag is based on the Norton design as opposed to the Win95 version which
was not as accurate and efficient).

If what you say is true and there is no way to lose data now during a defrag
operation, thank you for the information (always willing to teach this old
dog new tricks :-) that is one reason I subscribe to over 4000 emails a
month on mailing lists) and can you point me to the documentation stating
such.

Finally, we still do not agree. I did 100s of xcopy before you sent me the
email (several months ago) with the URL describing the POSSIBILITY of a
problem (thank you for the information) and never had a problem. I still say
calling it "unsafe" from my experience is a Chicken Little approach to the
"sky is falling." Stating the facts is stating facts but crying "the sky is
falling" from an opinion may be harmful. State the facts and let the user
decide if it is "safe" or "unsafe" for him/her self.

Can we agree with letting the user decide if it is safe/unsafe for him/her
without putting our own opinion on them? As with defrag/xcopy and everything
else in this world...nothing is "safe". I would rather give factorial
information and let the user decide. I give factual information as best I
can based on the most recent information I have been given and can
experience and check out other wise. Then, as I said, let the user decide.
Or in other other words, when it comes to opinions, they are like A-holes.
We all have one and my stinks as much as another ones sometimes. One
word...bode (sp?) <G>

Daniel Wysocki
Twin*.*Star Computers
770-498-2582 /800-816-0663
[log in to unmask]
Fast - Reliable - Wallet Friendly
-----Original Message-----
From: Earl Truss <[log in to unmask]>
To: PCBUILD - Personal Computer Hardware discussion List
<[log in to unmask]>
Date: Monday, June 07, 1999 11:55 AM
Subject: Re: [PCBUILD] Cannot get new hard drive to load Win98


>At 08:45 PM 6/5/99 -0400, you wrote:
>>If you have a power failure/fluctuation which causes you computer to
>>freeze, reboot, etc. you have nearly 100% guaranteed lost data since
>>when defragging, the program removes data from one area, saves it in
>>volatile RAM and then copies it to another area of the hard drive.
>>Until all the info in the RAM is copied to the hard disk, it is only in
>>RAM which lost forever if the puter freezes and or reboots. Lost data
>
>The above scenario is exactly why a defrag program does NOT work as you
>have described.  It reads the data from your disk, writes it to another
>area of the disk and THEN, once the data is safely transferred, the FAT is
>changed to remove the data from the old position.  There is no removal of
>data from one part of the disk while it is not stored in another part.  If
>a power failure or other hangup occurs at any point, the FAT will always
>point to valid data - either the original data or where the data has been
>moved.  If a defrag program actually worked as you have described, the
>window during which the disk could be damaged would be much too large and
>it would be too unsafe to use.
>
>>After several 100 xcopies, I have never had a problem except as
>>mentioned with a bat file. Remember, just because you have truncated
>>pointers (to include shortcuts), does not mean you will have a problem,
>>only if the xcopy changes it which it may not depending on the order of
>>copy.
>>
>Of course.  This is exactly the point.  It may be fine for you or someone
>who understands the problem to use XCOPY and correct any problems
>afterwards.  It is another thing to recommend this method to an
>inexperienced user who will not recognize the problem or know how to
>correct it without mentioning the possibility of the problem happening.  I
>think that is all that anyone here is saying.  We are not saying that this
>problem makes this method unsafe in all circumstances and that you WILL
>have problems, only that the instructions must include the warning of the
>possibility of problems, how to recognize the problem and how to correct
>it.  Of course, for inexperienced users, this should make other more safe
>alternatives more attractive.  Can we agree on this?
>
>

         The PCBUILD web site always needs good submissions.  If
          you would like to contribute to the website, send any
               hardware tech tips or hardware reviews to:
                           [log in to unmask]