Dealing with a Document of Death in Word 2003

[Guest]

I have an MSWord doc that, when opening, brings the CPU up to 100%, and stays
there till I kill word. It's all very well having crash handling -like the
dumprep uploader, the reload open files and the "remember which apps crashed
word on loading", but how do I deal with the problem?

Is that document officially deceased, never to be used again? Is there any
support mechanism for people with whom word toasts, like an email address to
mail the doc to see if it can be salvaged?

 

[Guest]

It may be worth trying File, Open and change Files of Type to "Recover Text
from Any File" and see if the text is recoverable that way.

 

[Margaret Aldis]

Better to follow these first:

http://www.word.mvps.org/faqs/apperrors/CorruptDoc.htm

http://www.gmayor.com/what_to_do_when_word_crashes.htm

 

[Daiya Mitchell]

Some tips for recovering corrupt documents here:

The first way to check for a corrupt document is to
copy the entire thing, *excluding* the last paragraph mark, into a new
document. That last paragraph mark holds a lot of information which can get
corrupted, and copying the text into a document with a fresh one keeps your
formatting, but can fix some glitches.

A paragraph mark is a gray ¶. Click on ¶ on the standard toolbar to show
nonprinting characters, including paragraph marks.

See this link for further info:

http://www.mvps.org/word/FAQs/AppErrors/CorruptDoc.htm

 

[Guest]

Some tips for recovering corrupt documents here:

The first way to check for a corrupt document is to
copy the entire thing, *excluding* the last paragraph mark, into a new
document. That last paragraph mark holds a lot of information which can get
corrupted, and copying the text into a document with a fresh one keeps your
formatting, but can fix some glitches.

A paragraph mark is a gray ¶. Click on ¶ on the standard toolbar to show
nonprinting characters, including paragraph marks.

See this link for further info:

http://www.mvps.org/word/FAQs/AppErrors/CorruptDoc.htm

I'm going to try all of these. The save as HTML looks interesting, and makes
me think if I could save as XML I could open it up in an IDE and patch by
hand, but that seems, well, dangerous. At the same time, saving to HTML will
lose interesting stuff.

I now suspect it may be a funny with one specific machine. I managed to open
up the file on another system, accepted all changes, saved as to a different
filename to get rid of stuff in the OLE filesystem. But still it crashes.
hmmm. Time to erase that vmware image, perhaps.

 

[Guest]

I had the same identical problem and the following solution that I just
discovered seems to be working. My document has automatic table of contents,
table of figures etc. (the kind you make by "Insert > Index and Tables..."),
and updating the numbering of those tables solved the CPU freak-out for me.

Do this by:

1. Select the entire document.
2. right-click anywhere in any table of contents, table of figures, etc
3. Choose 'Update Field' > 'update entire table' > OK
4. If you have more than one table of contents/fgures/etc you'll have to
click OK a few more times (I think, more than just once for each table,
though that doesn't seem to make sense).

This solution is not mentioned at
http://www.word.mvps.org/faqs/apperrors/CorruptDoc.htm

although that page does mention that corruption can be caused by automatic
list numbering. That gave me the idea to update the numbering - even though I
normally do that fairly anyways, I guess I hadn't for a while on that
particular document.

I've never figured out if you really need to select the entire document, nor
the difference between 'update page numbers only' and 'update entire table'.
I just figure, I want everything updated, always,

 

[Guest]

"finisterre" wrote:

Oops, I meant:

 

[Guest]

Retraction:

Previously I said that updating the fields of the automatically-generated
Table of Contents and Table of Figures solved the problem of a particular
document causing CPU usage to stay at 100%.

However, an hour or less after I doing that, the problem returned, so that
did not turn out to be a solution to the problem.

Sorry - shoulda waited a bit before posting.