Word 2004 taking 30% CPU even when idle w/ no doc open

[qiaob...@yahoo.com]

After recent upgrade of my PowerBook G4 to Tiger (10.4.10), I've
notice that Word 2004 constantly takes up 30+ % CPU even when it's
idle with no open document. I tried to re-install the entire Office
2004 for Mac but the problem remains the same. I never saw this
behavior when I was Word 2004 running Panther.

Any suggestion to a fix will be appreciated.

-QB

[rbfraven]

I wish I knew a fix. I first noticed MS Word 2004 using excessive
processor time on my 500Mhz G4 Powerbook a few days after I applied
the latest iTunes and Quicktime update (7/14) and the Microsoft update
11.3.6 (7/13). Prior to that I upgraded to 10.4.10 (6/21) from 10.4.9
but the problem seemed to occur after the previously mentioned
updates. These are the only system changes I am aware of before the
problem occurred.

After those changes, the fan started coming on more frequently, and
Unix top and Activity Monitor showed that Word was always using 50%
even when NO files were open. When I quit Word, CPU usage decreases by
50%. When I start it, CPU usage increases by 50%. My temporary
circumvention is to quit Word immediately after using it. I used to
always run with Word as a running application.

I have re-installed all of the previously mentioned updates but the
problem remains.

This problem is also described in the post "7.2 Install affected Word".

[Elliott Roper]

In article <1187496342....@a39g2000hsc.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

> I wish I knew a fix. I first noticed MS Word 2004 using excessive
> processor time on my 500Mhz G4 Powerbook a few days after I applied
> the latest iTunes and Quicktime update (7/14) and the Microsoft update
> 11.3.6 (7/13). Prior to that I upgraded to 10.4.10 (6/21) from 10.4.9
> but the problem seemed to occur after the previously mentioned
> updates. These are the only system changes I am aware of before the
> problem occurred.

I see you posted this already without getting a useful answer. My
suggestion of following the recipe in Apple's KB 306043 would have been
useless, since it was only supposed to be for Intel Macs, and I somehow
missed the top line of your post. Sorry.

You could try some geeky diagnostic testing by sampling Word with
Activity monitor. The sampler will dive into Word 300 times and report
where in Word all the fun was happening. It might be interesting to
report back your results, although it is unclear whether anyone outside
Microsoft could usefully interpret the them, since older "CFM" software
such as Word is rather coy about naming the procedures sampled.

>
> After those changes, the fan started coming on more frequently, and
> Unix top and Activity Monitor showed that Word was always using 50%
> even when NO files were open. When I quit Word, CPU usage decreases by
> 50%. When I start it, CPU usage increases by 50%. My temporary
> circumvention is to quit Word immediately after using it. I used to
> always run with Word as a running application.
>
> I have re-installed all of the previously mentioned updates but the
> problem remains.

Re-installing Word is unlikely to help. You might try repairing
permissions with Disk Utility. Although I think it is about as useful
as a chocolate teapot, it is easy and safe and gives you a warm
feeling. If it does do anything useful, please report back and I'll eat
my words. I'd be inclined to download and run the 10.4.10 combo
updater from Apple, possibly with added forced rebinding as described
in Apple knowledge base article 306043 even though it was directed at
Intel Macs. It would be a sort of blanket fix for things that went
missing in Apple's Software Update.

--
To de-mung my e-mail address:- fsnospam$elliott$$
PGP Fingerprint: 1A96 3CF7 637F 896B C810 E199 7E5C A9E4 8E59 E248

[rbfraven]

I did read the 306043 article that you mentioned and I did re-install
10.4.10 using the combo update in safe mode. In fact I re-installed QT
7.2, Office 11.3.6, and the combo 10.4.10 in safe mode. I almost re-
installed Office but was unsure if I could preserve all of my existing
Entourage calendar data. QB, who started this thread, did re-install
Office 2004 without any effect, so I don't think I will do it. As for
repairing permissions, that happens every few days automatically when
I run SuperDuper to make a backup. I tend to do that anyway before and
after major software updates. I did not do the rebinding that is
recommended in 306043 because I was unsure about doing it for a PPC as
it was recommended for an Intel machine. I don't know if it could have
serious side-effects or if it is a safe operation to do at anytime. I
will try the Activity Monitor sampling that you mentioned. If we knew
what Word is doing, that might suggest a solution. Word is the only
app I have found with this problem on my machine.

Thanks for the follow-up.

Bob Fowles

[Elliott Roper]

In article <1187525736....@a39g2000hsc.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

I think the pre-binding is pretty benign. It is what happens when Mac
updates say "Optimizing System Performance"

You have up-to-date clones. Go for it!

[John McGhie]

I don't think Word is actually "doing" anything :-)

Word (like many applications) is an "idle loop processor". It does ALL of
its processing in the idle loop (makes it a good multi-tasking citizen).

However, this can confuse Activity Monitor. If you have nothing much
running, Word's foreground loop may show that it is using 30 per cent, but
what that really means is "30 per cent of the CPU being used", which is
quite different from "30 per cent of the available CPU".

All Word is doing is polling the keyboard buffer ‹ "Any keystrokes for me?
Nup! OK, back to sleep!" There's actually nothing happening, but because
nobody else is doing anything, it represents a lot of the CPU being used.

The real test would be to start something you know is power-hungry (e.g. Rip
a few tunes with iTunes or a DVD...) Then have look and see if Word is
still showing up as 30 per cent.

Chances are it will drop back to around five per cent. Word is still using
the same amount, but now it's a fraction of a much larger amount of CPU now
being used :-)

Cheers

On 19/8/07 9:45 PM, in article
1187525736....@a39g2000hsc.googlegroups.com, "rbfraven"
<r...@psu.edu> wrote:

--
Don't wait for your answer, click here: http://www.word.mvps.org/

Please reply in the group. Please do NOT email me unless I ask you to.

John McGhie, Consultant Technical Writer
McGhie Information Engineering Pty Ltd
http://jgmcghie.fastmail.com.au/
Sydney, Australia. S33°53'34.20 E151°14'54.50
+61 4 1209 1410, mailto:jo...@mcghie.name

[Elliott Roper]

In article <C2EF9E48.6A4F%jo...@mcghie.name>, John McGhie
<jo...@mcghie.name> wrote:

> I don't think Word is actually "doing" anything :-)

In this case, that might be debatable.
1. Bob F's fans are kicking in.
2. Word does not normally do that. My Boko's Word 2004 shows 1.6% on
Activity Monitor when totally idle. OK, at 1 GHz, it is twice as quick
as Bob Fowles' 500 MHz, but that still means his copy guzzles 10 times
more cpu when idle. He does have something wrong.

>
> Word (like many applications) is an "idle loop processor". It does ALL of
> its processing in the idle loop (makes it a good multi-tasking citizen).
>
> However, this can confuse Activity Monitor. If you have nothing much
> running, Word's foreground loop may show that it is using 30 per cent, but
> what that really means is "30 per cent of the CPU being used", which is
> quite different from "30 per cent of the available CPU".

Err no. Activity Monitor *does* display the idle time. Its reported
CPU% really is a percentage of what each processor can do flat out.
(Bob's and my Bokos have only one)

It is instructive to sample Word with Activity Monitor.
You will see that there is no polling going on in an idle Word. All the
threads are sitting in mach_msg_trap or similar. That means nothing is
charged to the application until the OS hands it control when an event
that Word is waiting for actually occurs. OK, one of those events may
be a timer as some kind of catch-all, but that would be constant, small
and invariant with load. Sure, once it gets control, it might well
choose to run round like a pre-historic headless chicken, but it should
not get control till one of the events it has declared an interest in
actually fires.

If you look at a multi-cpu sample it will be different. There each
thread except the main one is sitting on a spinlock. No cpu charge to
Word. The main thread is as described above.

>
> All Word is doing is polling the keyboard buffer ‹ "Any keystrokes for me?
> Nup! OK, back to sleep!" There's actually nothing happening, but because
> nobody else is doing anything, it represents a lot of the CPU being used.

That's how OS9 and earlier did it. We're not in Kansas any more. Word
will stay asleep till an event for it fires.

> The real test would be to start something you know is power-hungry (e.g. Rip
> a few tunes with iTunes or a DVD...) Then have look and see if Word is
> still showing up as 30 per cent.
>
> Chances are it will drop back to around five per cent. Word is still using
> the same amount, but now it's a fraction of a much larger amount of CPU now
> being used :-)

There is a small element of that, once Word and the other processes are
potentially guzzling > 100% of what's available in user mode. The
ordinary user perceives that as Word going like a slug.

Bob's fans should not be kicking in.
Therefore either his OS is pooched and dealing out events to Word like
there is no tomorrow, or his Word is borked and every time it gets
control it is doing something time-consuming and pointless. A sample
would show the latter at work.

[Bill Tucker]

Are any of you using Endnote? If so, follow these instructions I
received from Endnote technical support. I did this, and CPU usage
with the re-formatted Word document is back to normal. However, after
I Format Bibliography, I see the CPU usage climb back up. So, I
followed these directions again to revert all the citations. Moral of
the story? If using Endnote with Office 2004 for Mac, don't format the
bibliography until the very end.

You may need to clean up the field codes in your document. Try the
following steps:

1. Make a backup of your document.

2. Choose the "Tools > EndNote > Unformat Citations" command, which
will remove the reference list and revert the citations.

3. Do a [Apple]+A on the keyboard to highlight everything.

4. Do a [Apple]+6 (above the "T" and "Y" key) to remove any
additional hidden field codes.

5. Do a [Apple]+C to copy the highlighted text.

6. Open a new document and do a [Apple]+V to paste.

You should now be able to format this cleaned up version of the
document without running into problems.

[qiaob...@yahoo.com]

Thanks for the suggestions. Here are what I found:

1) The Word problem is taking away real CPU cycles (not just idle
cycles).

- When Word is idle with no document open, and no other application
running: Word takes ~30% CPU
- When Word is idle with no document open, and start a mp4 video
converter: Word takes ~25% CPU, Video Converter takes ~58% CPU
- Quite Word but leave the video converter running: Video Converter
takes ~85% CPU

2) Here is the Activity Monitor sampling results (Word is PID 2942).
Any ideal?

2007-08-21 14:37:14.922 sample[2945] Waiting 20 msecs for System
framework to be loaded...
2007-08-21 14:37:14.945 sample[2945] Waiting 40 msecs for System
framework to be loaded...
2007-08-21 14:37:14.986 sample[2945] Waiting 80 msecs for System
framework to be loaded...
2007-08-21 14:37:15.070 sample[2945] Waiting 160 msecs for System
framework to be loaded...
2007-08-21 14:37:15.249 sample[2945] Waiting 320 msecs for System
framework to be loaded...
2007-08-21 14:37:15.618 sample[2945] Waiting 640 msecs for System
framework to be loaded...
2007-08-21 14:37:16.292 sample[2945] Waiting 1280 msecs for System
framework to be loaded...
2007-08-21 14:37:17.655 sample[2945] Waiting 2560 msecs for System
framework to be loaded...
2007-08-21 14:37:20.219 sample[2945] *** Process unreadable: 1
Analysis of sampling pid 2942 every 10.000000 milliseconds
Call graph:
267 Thread_0f03
267 0x5820078
267 0x686c464
267 0x58205fc
267 0x58f3a9c
267 0x4c93820
267 0x238d154
267 0x22881b4
267 0x2288b20
267 0x167d4ac
188 0x167dba8
188 0x110e348
188 0x110e348
75 0x167def8
75 0x1691578
75 0x4c93b3c
74 0x5904074
74 0x59041e4
74 0x5909ffc
74 0x590b414
74 EventAvail
74 GetNextEventMatchingMask
74 0x2288b20
42 0x167d4ac
18 0x167da0c
18 0x167d6b8
18 0x167ba5c
18 0x16696dc
18 0x1173994
18 dladdr
10
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
10
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
6 dladdr
2
ImageLoaderMachO::getExportedSymbolAddress(ImageLoader::Symbol const*)
const
2
ImageLoaderMachO::getExportedSymbolAddress(ImageLoader::Symbol const*)
const
12 0x167da1c
12 0x167d6b8
12 0x167ba5c
12 0x16696dc
12 0x1173994
12 dladdr
10 dladdr
2
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
2
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
11 0x167da2c
11 0x167e3f8
11 0x167ba5c
10 0x16696dc
10 0x1173994
10 dladdr
8 dladdr
1
ImageLoaderMachO::getExportedSymbolAddress(ImageLoader::Symbol const*)
const
1
ImageLoaderMachO::getExportedSymbolAddress(ImageLoader::Symbol const*)
const
1
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
1
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
1 0x167ba5c
1 0x167dba8
1 0x110e348
1 0x110e348
18 0x167d494
10 0x167d6b8
10 0x167ba5c
10 0x16696dc
10 0x1173994
10 dladdr
6 dladdr
3
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
3
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
1
ImageLoaderMachO::getExportedSymbolAddress(ImageLoader::Symbol const*)
const
1
ImageLoaderMachO::getExportedSymbolAddress(ImageLoader::Symbol const*)
const
8 0x167d710
8 0x167801c
8 0x16696dc
8 0x1173994
8 dladdr
6
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
6
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
2 dladdr
13 0x167d4c0
13 0x167d6b8
13 0x167ba5c
13 0x16696dc
13 0x1173994
13 dladdr
11 dladdr
2
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
2
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
1 0x167d3f8
1 0x167b218
1 0x167bc54
1 0x167bc54
1 0x5904058
1 0x59040fc
1 0x590b414
1 EventAvail
1 GetNextEventMatchingMask
1 0x2288b20
1 0x167d4ac
1 0x167da2c
1 0x167e3f8
1 0x167ba5c
1 0x16696dc
1 0x1173994
1 dladdr
1 dladdr
1 0x167da54
1 0x167d710
1 0x167801c
1 0x16696dc
1 0x1173994
1 dladdr
1 dladdr
1 0x167dd2c
1 0x167d6b8
1 0x1699124
1 0x1699124
1 0x167de4c
1 0x168b5a4
1 0x167ba5c
1 0x16696dc
1 0x1173994
1 dladdr
1
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
1
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
1 0x167dfd0
1 0x167d5f0
1 0x167d5f0
267 Thread_1003
267 0x112ed08
267 MerpUnregisterCFMFragment
267 0x110e348
267 0x110e348

Total number in stack (recursive counted multiple, when >=5):
9 0x1173994
9 0x16696dc
9 dladdr
7 0x167ba5c
7 ImageLoaderMachO::getIndexedExportedSymbol(unsigned)
const
5 0x167d6b8

Sort by top of stack, same collapsed (when >= 5):
0x110e348 456
dladdr 45
ImageLoaderMachO::getIndexedExportedSymbol(unsigned)
const 25
Sample analysis of process 2942 written to file /dev/stdout
Sampling process 2942 each 10 msecs 300 times

[Elliott Roper]

In article <1187725602....@i13g2000prf.googlegroups.com>,
<qiaob...@yahoo.com> wrote:

> Thanks for the suggestions. Here are what I found:
>
> 1) The Word problem is taking away real CPU cycles (not just idle
> cycles).
>
> - When Word is idle with no document open, and no other application
> running: Word takes ~30% CPU
> - When Word is idle with no document open, and start a mp4 video
> converter: Word takes ~25% CPU, Video Converter takes ~58% CPU
> - Quite Word but leave the video converter running: Video Converter
> takes ~85% CPU
>
> 2) Here is the Activity Monitor sampling results (Word is PID 2942).
> Any ideal?

Well, whatever is calling
ImageLoaderMachO::getIndexedExportedSymbol(unsigned) const
is whacking your machine to death.
Are you running EndNote or some other add-on to the basic Office
product?

[rbfraven]

I tried the sampling also and the results are similar to QB's. I also
saw a couple of spin_locks in the sampling. When I start word, it
takes 50% CPU. If I then start apple.com/trailers and click on Bourne
Ultimatum, Word drops to <20% and %idle is <2% (according to unix
top). Again, this is the only app I have found with this problem. I
would do the upate_prebinding command but I am in the middle of some
projects and don't want to risk any further problems. My PB firewire
port is dead and I wouldn't be able to SuperDuper clone back to my HD
in case of a poblem (I backup up via a PC card with a firewire port
but I can't boot from my external clone unless attached to the
firewire port). I've never run Endnote.

Bob Fowles (mac user since 1988)
rbf
AT
psu.edu

[Elliott Roper]

In article <1187737836.0...@i13g2000prf.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

That sounds like a plan.
After all those negative results, and if the prebinding is no help, you
are up for an archive and install of OS X. So get your projects safe
first. Archive and install is quite easy, but you don't want to take
too many risks for a bit of fan noise while you can't restore backups.

[qiaob...@yahoo.com]

I never used EndNote. The Mac OS and Office 2004 on my Powerbook were
recently freshly installed (just upgraded to Tiger) and both were
updated with whatever available Apple and Microsoft updates. No add-
ons.

[rbfraven]

After you installed 10.4.10, did you also install Quicktime 7.2 and
Office update 11.3.6? I installed 10.4.10 on 6/21 but did not notice
the Word problem until after I installed QT 7.2 and Office 11.3.6. It
could possibly be that it started after installing 10.4.10 and that I
didn't really pay attention to it (fan coming on) until after the
other two updates, but to the best of my recollection it started after
7.2 and 11.3.6.

[Elliott Roper]

In article <1187790676.8...@x40g2000prg.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

If either of you have an Intel Mac and the problems started happening
after the QT 7.2 install, then a OS X 10.4.10 combo updater, followed
by the prebinding exercise if needed is a really good idea.

[rbfraven]

On Aug 22, 10:08 am, Elliott Roper <nos...@yrl.co.uk> wrote:
> If either of you have an Intel Mac and the problems started happening
> after the QT 7.2 install, then a OS X 10.4.10 combo updater, followed
> by the prebinding exercise if needed is a really good idea.
>

I have a 500Mhz Titanium G4 PB and QB in his original post stated he
has a G4 PB.

Bob Fowles

[Elliott Roper]

In article <1187792031.1...@l22g2000prc.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

In that case, it probably won't hurt, but will probably be no help
either.

The only thing left I can offer you is that my fully updated G4 PB does
not exhibit that behaviour. It uses 1.6% cpu when idle.

You may possibly infer from that, that a complete re-installation of
operating system and Office and all their updates and uncles and aunts,
*might* get 'em into the same happy state that mine is in.

If you want company in your misery, I forgot to mention that this
self-same Powerbook would not boot after the original 10.4.10 update
and I took the opportunity to have a giant clean-up, reinstalling OSX
10.1.10 and all my applications.

I'm sorry I could not help you more productively. Please post back with
whatever magic spell got you working coolly again.

[rbfraven]

Elliott,

Thanks very much for your consideration and suggestions. I will
continue living with the Word problem until I find a fix, or a future
update fixes it, or someone posts a fix, or until I replace my aging
G4 PB with a MBP in a few months. If I do fix it, I will post how I
did it.

Bob Fowles

[Elliott Roper]

In article <1187833856.1...@l22g2000prc.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

Thank you. Somewhere in a forgotten recess of my brain I have a
recollection of something similar going wrong here a long time ago. It
is annoying that I can't remember what made it go away.

[John McGhie]

Hi Elliott:

The upgrade to 2004 may have done it.

They changed the handling of threading in 2004 to get around Word X's high
background noise.

There's something in "my" memory about the mouse driver having been involved
in this. Elliott, it wouldn't affect you, we presume...

But if either of the other posters are using a high-precision mouse, it may
be interesting to unplug it and see what effect that has. The old version
in X could get bogged down processing an unending steam of mouse messages.

Cheers

On 23/8/07 7:32 PM, in article 230820071102550958%nos...@yrl.co.uk, "Elliott
Roper" <nos...@yrl.co.uk> wrote:

> In article <1187833856.1...@l22g2000prc.googlegroups.com>,
> rbfraven <r...@psu.edu> wrote:
>
>> Elliott,
>>
>> Thanks very much for your consideration and suggestions. I will
>> continue living with the Word problem until I find a fix, or a future
>> update fixes it, or someone posts a fix, or until I replace my aging
>> G4 PB with a MBP in a few months. If I do fix it, I will post how I
>> did it.
>
> Thank you. Somewhere in a forgotten recess of my brain I have a
> recollection of something similar going wrong here a long time ago. It
> is annoying that I can't remember what made it go away.

--

[rbfraven]

Just an update. I just discovered that my Dreamweaver MX 2004 (also
with no open files) also has the same CPU consumption problem. It is
now taking more than Word 2004 - 60-70%. Unix top shows most CPU is
User and not System for both Word and Dreamweaver. My fan comes on
soon after either one is started. It is good bet that the same cause
is affecting both pieces of software. It is possible that the cause is
hardware (an aging Ti book) but the fact that it started only after
recent Apple and Word updates and that other G4 PBs are seeing this
also (at least for word) leads me to suspect one or a combination of
the software updates (10.4.10, QT 7.2 and Office 11.3.6). I will
concentrate now on pursuing this as a 10.4.10 problem.

Bob Fowles

[qiaob...@yahoo.com]

I discovered a simple work-around: just open any pull-down menu in the
Word (e.g., click Word->About Word and leave it open) and this will
stop whatever stupid thing Word is doing in the background. This way
you don't need to close all your docs and quit Word in order to get
your CPU back.

On Aug 24, 8:30 pm, rbfraven <r...@psu.edu> wrote:
> Just an update. I just discovered that my Dreamweaver MX 2004 (also

> with no open files) also has the sameCPUconsumption problem. It is
> now taking more thanWord2004 - 60-70%. Unix top shows mostCPUis
> User and not System for bothWordand Dreamweaver. My fan comes on

> soon after either one is started. It is good bet that the same cause
> is affecting both pieces of software. It is possible that the cause is
> hardware (an aging Ti book) but the fact that it started only after

> recent Apple andWordupdates and that other G4 PBs are seeing this
> also (at least forword) leads me to suspect one or a combination of

[rbfraven]

That is very interesting and works just as you say.

I had posted previously that DreamWeaver was also consuming excessive
CPU. That is only true while it is waiting for the first file to be
opened. After that CPU usage is normal (low). So Word 2004 is the only
app I have found that always uses 50% CPU with or without open files.

John McGhie suggested disconnecting any "high-precision" mouse. I have
an optical IntelliMouse and I tried disconnecting it but that made no
difference.

I tried the update_prebinding mentioned previously but that made no
difference.

The next thing to try is an Archive-Install. I wouldn't be surprised
if that fixes it as I have never done a clean install since I got my
G4 500Mhz PB in Spring 2001. Who knows what cruft is floating around
under the hood after installing all updates since 10.1 plus dozens of
apps and their updates.

In my case, I will just tolerate it until I upgrade to a MBP with
Leopard.

Thanks for the update.

Bob Fowles

[rbfraven]

I solved my "Word 2004 Excessive CPU Problem" with the help of
http://groups.google.com/group/microsoft.public.mac.office/browse_thread/thread/d9bc05d5f147fe55/cc8809032f4c250f?lnk=gst&q=renaud&rnum=2#cc8809032f4c250f
and especially the blog http://mactip.blogspot.com/2007/07/dependent-dylib-is-not-prebound.html
which it references. I had previously run update_prebinding but failed
to notice an error message. Apparently that error prevented
update_prebinding (UP) from completing (see link). UP uses the data
file /var/db/dyld/update-prebinding-paths.txt which is a list of files
to be operated on. By using -debug on UP you can see exactly where UP
stops. In my case it was failing on a group of files in /usr/X11R6/lib
and not doing any after those. I made a copy of the file and with
emacs edited those lines by inserting "#" at the beginning to comment
out those filenames. After re-running UP, MS Word now uses less than
3% which is quite a reduction from the 50% it was using. My thanks to
Renaud Blanch (1st link above), the blog author (2nd link above) and
anyone else who responded to my earlier posts about this problem.

I'm grateful to have this solved after 2.5 months but several
questions remain including (1) Which update 10.4.10, QT 7.2 or MS
Office 11.3.6 caused the problem?, (2) How does this file get updated,
and (3) Why was it in error?

Bob Fowles

[Elliott Roper]

In article <1190939462....@22g2000hsm.googlegroups.com>,
rbfraven <r...@psu.edu> wrote:

> I solved my "Word 2004 Excessive CPU Problem" with the help of

<snip>

> I'm grateful to have this solved after 2.5 months but several
> questions remain including (1) Which update 10.4.10, QT 7.2 or MS
> Office 11.3.6 caused the problem?, (2) How does this file get updated,
> and (3) Why was it in error?

From reading the article in Apple's KB it looked as though the QT 7.2
update broke (probably installed a wrong version of) some library
needed by the Rosetta PPC emulation. It is strange that the problem did
not show on all machines or all Rosetta programs, and for those who
were affected, how it affected them in different ways, from a complete
non-start to your cpu hogging. QT 7.2 became available about the same
time as Office 10.3.6 so there remains a slight possibility that the
order those updates were applied was relevant. I fervently hope not.
The two should have been utterly disjoint.

I did QT 7.2 last and it did not bite me, although I later did the
combo and pre-binding steps when all the problem variations became
known. I never saw anything wrong here. I just did it for safety.

[John McGhie]

Hi Bob:

Many thanks to YOU for getting back to tell us what your solution was.

That's a particularly nasty bug to find...

Cheers

On 28/09/07 10:01 AM, in article
1190939462....@22g2000hsm.googlegroups.com, "rbfraven"
<r...@psu.edu> wrote:

--

Don't wait for your answer, click here: http://www.word.mvps.org/

Please reply in the group. Please do NOT email me unless I ask you to.

John McGhie, Consultant Technical Writer
McGhie Information Engineering Pty Ltd
http://jgmcghie.fastmail.com.au/

Nhulunbuy, Northern Territory, Australia