Saves my Day, Again

by Jonathan Gennick

Awhile back I wrote about my
experience installing for my family
, pointing out that it's just
the right introduction to open source for many people. I said then that
saved my day, because it gave me an office suite I could freely install for
my children. This week saved my day again, by recovering the contents
of a corrupt, but very critical, Microsoft Word file.

It was Wednesday morning, 30 April. I was several pages into editing a chapter
written in Microsoft Word when things took a bad turn. Word got very sluggish,
and then "locked up". Windows Task Manager showed winword.exe
monopolizing 99% of my CPU. The CPU (on my Thinkpad)
got hot enough that the fan kicked in. Finally, Word died of some fatal error,
dutifully reported the error to home base in Redmond, restarted itself, and
"recovered" my file. Except Word didn't recover my file, not really.
The problem began to repeat itself the moment I began typing again. I suspected
some sort of internal corruption with my document file.

After several of Word's error, restart, recover cycles my heart began to sink.
I was not keen to lose the file with all the edits I'd made so far. I went to
Microsoft's web site and spent about 45 minutes downloading and installing all
the available Microsoft Word updates, but to no avail. I tried saving to a new
filename, but the problem persisted. I created a new, blank, Word file, and
copied the chapter text from the problem file to the new file, but the corruption
was copied too. I opened and saved the file using Word 2000, which I believe
to be more stable than Word XP, but still the problem lived on.

Finally I got to thinking that perhaps Word was just too darn good at copying
it's own files, and that I needed to pass my file through something other than
Word in order to filter out the corruption. Opening the problem file in,
I saw that the file contents looked intact, so I saved the file as an,
sxw file. Then I opened that sxw file using, and saved
the file again, this time as a Word file. With a bit of trepidation, I then
opened the new Word file using Word and began typing. Everything worked! All
my previous edits were intact. All the text was intact, including revision marks
and embedded comments. The corruption was gone. Happy, and very relieved, I
went to lunch.

I know that is not perfect—it has its own problems—but
I find it very ironic that Word couldn't fix it's own file, that the only way
I could find to save my Microsoft Word document was to filter it through a competing
office suite, and an open source office suite at that. And this despite the
fact that Microsoft has apparently gone to great lengths to add document recovery
features to Word. My kudos to the team for developing a product
robust enough to save my day, again.

Speaking of, I've been editing three books for which all the writing
and editing has been done using The first of these books, C++
In A Nutshell
by Ray Lischner,
just went to print Tuesday this week. The other two titles,
which are moving through the production process now, are Essential
by Jennifer Vesperman, and UML Pocket Reference by Dan Pilone.


2003-05-02 09:22:29
This is a fabulous example of the value gained when we move off of a monoculture. Word and OO, by virture of being different implementations, are not vulnerable to the same problems.
2003-05-02 09:51:34
Word file rescues
I've also used for this purpose. A couple of other tips for problem Word files:

- In Word, try saving to a non-DOC format file, like RTF, then open that file

- Keep a copy of AbiWord around to help rescue Word files in a similar fashion to the technique used with I've had AbiWord work when or Word tricks wouldn't work

2003-05-02 10:44:43
OpenOffice saved my bacon with an Excel file
Just last week, I had a similar problem with an Excel file. The data on the spreadsheet wasn't important, but the contents of the VBA macro I was writing would've taken me several hours to recreate. When I saved the spreadsheet at the end of the day (after having save multiple times throughout the day), I got a message saying that "Excel.exe has generated errors" and it quit. I rebooted, opened the spreadsheet, got the same error.

After about a half hour of poking around Microsoft's website it dawned on me that I could try opening it in OpenOffice. I opened it, did a "save as" and saved it to Excel format under a different name. When I re-opened it in Excel, some of the formatting was lost, but all my VBA code was there with all my changes, including the ones from the last save that killed Excel. I was very happy.

2003-05-03 06:07:41
Word 97 couldn't open a Word 97 file
I recall about five or so years ago when I prepared a resume with Word 97. I was getting asistance from a professional on resume preparation, and this was meant to be something that she could review. Well, _her_ copy of Word 97 couldn't even open the file.

I know the floppy was OK, because I took it back home, and I was again able to look at the document on _my_ machine.

This sort of blew me away...Word 97 not being able to even open a Word 97 document.

In the final analysis, going along with previous comments, you can just do some hard work and open a Word file with a text editor or a hex editor, and start looking through all the bloat for your critical text.

2003-05-03 14:09:21
and the sxw file is half the size of the doc ;)
Had the same experience ... and the sxw file is half the size (and sometimes even smaller) of the WORD doc file ;)
2003-05-03 19:03:10
So why is everyone still using Mircosoft word.
Jonathan Gennick
2003-05-04 06:20:12
I ask myself that question sometimes. There are some reasons I still use Word, and it'd be worth thinking those through systematically. Look for another weblog entry on this topic, later in the week.
2003-05-04 06:58:08
and the sxw file is half the size of the doc ;)
If you using MS OS, please change *.sxw to *.zip or *.rar and open it.
2003-05-05 00:17:34
Several reasons, some good some less so:
1) you have it already and it works for you so why change to something you have to learn all over again?
2) corporate policy. Many (most?) companies, especially large ones, don't want to use any product if there's no company behind it that can provide support (or if push comes to shove be sued for damages).
3) many people have never heard of Open Office or tried it at a time it was very poor and gave up on it (I was disappointed when I tried it, couldn't compare to MS Office which I have on a company license...).
4) distrust of open source. Too many OS projects go wrong or are so full of bugs as to be useless.
2003-05-05 03:31:09
OO calc saved my day
Same problem with an excel sheet.
Just opened it in OpenOffice Calc, and told my users to start installing this wonderful office suite.

They don't use it 100%, but we're slowly moving onto a clean, GPL office suite.

Thomas Mathiesen

2003-05-05 05:29:13
Last year I had a similar problem with Powerpoint; it was unable to open a presentation produced by itself.
Impress used as a filter did the job without problems.
2003-05-05 05:45:34
"2) corporate policy. Many (most?) companies, especially large ones, don't want to use any product if there's no company behind it that can provide support (or if push comes to shove be sued for damages)."

I don't know if you've ever read the EULA that you have to agree with to use Word (or any proprietary software) but it specifically states that Microsoft will NOT be held liable for anything. So, you CAN'T sue them for damages.

As for support, I have never found Microsoft to provide good support for any product.

2003-05-06 02:54:10
Star Office 5.2 saved my bacon many times
during my final year at university, I was trying out StarOffice 5.2 on my pc.A friend had corrupted a word 97 document, the night before the work was due in, so I thought I would try opening it in StarOffice. It opened fine other than a few formatting issues. Everyone was amazed. Since then I've always had a copy of StarOffice or OpenOffice on my PC.
2003-05-06 11:12:35
Another rescue trick
Another trick to rescue a corrupt file in MS Word is to turn on the paragraph marks with the toolbar, select all the document *except for the last paragraph mark*, and then copy and paste into a new document. A lot of information (so usually a lot of the corruption) is stored in that last paragraph mark.

Random Helpdesk Analyst and MS Word expert :)

2003-05-09 07:47:17
Um...Open Office?
As an O'Reilly contributor catering to a heavily geek crowd, shouldn't he have at least taken the time, before writing this article, to realize that Open Office is a trademarked name and that the office suite that is the basis for Sun's StarOffce is actually called

2003-05-13 01:08:35
Had also the same problem, same fix.
With Win Word 97 I had a similar problem, The document suddenly started to crash any MSWord which tried to open it. The way through OO saved me the info.
Thanks OO guys.
Jonathan Gennick
2003-06-01 17:23:40
Um...Open Office?
I corrected the article, and also my previous article. I must say though, that going forward this ".org" business will be a losing battle. The words "OpenOffice dot org" are just too much of a mouthful for the average man-on-the-street to want to bother with.
2004-12-16 00:52:34
OpenOffice saved my bacon with an Excel file
Could you tell what open office is and where to get it?
Jonathan Gennick
2004-12-16 05:47:41
OpenOffice saved my bacon with an Excel file
OpenOffice is an office suite similar in concept to Microsoft Word. You get word-processing, spreadsheeting, and I forget what all else. You can learn more by visiting
2005-02-01 18:15:15
No, you're not.
To be precise, is LGPL or SISSL, depending on what you want to use it under, but it's not GPL.
2005-03-17 23:21:41
Very often Sun StarOffice open files Ms Office and OpenOffice can't.

Dave Hall
FDRLab Data Recovery Centre Consultant