Page file setting?

[Ron Martell]

Tue, 14 Feb 2006 15:15:35 -0000 from Gerry Cornell


Is there no procedure to correct an erroneous community opinion? That
would be worse even than Wikipedia.

I know that nothing should surprise me at Microsoft, but I'd think
they'd want to correct an error that's published under their logo,
disclaimer or no disclaimer.

I once spent three years trying to get an incorrect KB article
regarding Windows 95/98/Me corrected. It was eventually changed, but
still contained several glaring errors.

"The only good thing about beating your head against a brick wall is
that it sure feels good when you stop."

Ron Martell Duncan B.C. Canada
--
Microsoft MVP (1997 - 2006)
On-Line Help Computer Service
http://onlinehelp.bc.ca

"Anyone who thinks that they are too small to make a difference
has never been in bed with a mosquito."

 

[Gerry Cornell]

Stan

This Article, which can be criticised for the same reason, was formulated
by Microsoft and no doubt has been reviewed many times!

How to configure paging files for optimization and recovery in Windows XP
http://support.microsoft.com/kb/314482/en-us

--

Hope this helps.

Gerry
~~~~
FCA
Stourport, England

Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~

 

[Ken Blake, MVP]

Tue, 14 Feb 2006 15:15:35 -0000 from Gerry Cornell


Is there no procedure to correct an erroneous community opinion? That
would be worse even than Wikipedia.

I know that nothing should surprise me at Microsoft, but I'd think
they'd want to correct an error that's published under their logo,
disclaimer or no disclaimer.

Personally, although I agree completely with Ron Martell, I don't think this
KB article is as wrong as all that. It clearly doesn't say the page file
*should be* 1.5 times the amount of RAM. With reference to how big the page
file should be, the article says "There is no single answer to this
question, because it depends on the amount of installed RAM and how much
virtual memory that workload requires. If there is no other information
available, the normal recommendation of 1.5 times the amount of RAM in the
computer is a good place to start."

"A good place to start," not "should be." Two very different statements. If
I were writing the article, I would write it somewhat differently, and talk
about how to get more information, rather than what to do in the absence of
information, but the statement there essentially says that if you don't have
any more information, leave the Windows default alone. I think that, as it
stands, that's a hard statement to argue with.

 

[Stan Brown]

Tue, 14 Feb 2006 12:18:17 -0800 from Ron Martell

I once spent three years trying to get an incorrect KB article
regarding Windows 95/98/Me corrected. It was eventually changed, but
still contained several glaring errors.

"The only good thing about beating your head against a brick wall is
that it sure feels good when you stop."

And in a similar vein: "Never try to teach a pig to sing. You'll only
waste your time and annoy the pig." -- Heinlein

It's first priority with me to correct errors on my site when they're
drawn to my attention. But I guess Microsoft can't afford the
resources to do that on its site.

Thanks for the info - it's always better to know than not to know.

 

[Stan Brown]

Tue, 14 Feb 2006 15:00:09 -0700 from Ken Blake, MVP

If there is no other information
available, the normal recommendation of 1.5 times the amount of RAM in the
computer is a good place to start."

"A good place to start," not "should be." Two very different statements

But both wrong. "1.5 times RAM" is _not_ a good place to start. For
Windows XP, _no_ setting is a good place to start. (And a multiple of
RAM, any multiple, was always stupid, as has been pointed out in this
very thread. Even on older Windows where it made (more) sense to set
the page file size, the desirable size was a decreasing function of
RAM size (disregarding what programs would be run).

I accept that this error's not going to get fixed, but let's not make
it out to be anything but what it is.

 

[Ken Blake, MVP]

Tue, 14 Feb 2006 15:00:09 -0700 from Ken Blake, MVP


But both wrong. "1.5 times RAM" is _not_ a good place to start. For
Windows XP, _no_ setting is a good place to start. (And a multiple of
RAM, any multiple, was always stupid, as has been pointed out in this
very thread.

I agree than any multiple of RAM makes no sense and I've often pointed out
the same here myself. My point is not that one should use any multiple of
RAM. It's that in the absence of any other information on which to base a
change to the Windows default setting, it's best to leave that default
alone, rather than run the risk of making it worse than it already is. I'm
not recommending a multiple of RAM, but it *happens* that that Windows
default is a multiple of RAM

My interpretation of what the article in question says is the point I made
above: if you don't know anything about the circumstances and what setting
is best, leave the default setting in place.

Even on older Windows where it made (more) sense to set
the page file size, the desirable size was a decreasing function of
RAM size (disregarding what programs would be run).

Absolutely. I've not only said the same thing here in the Windows XP
newsgroups, but also in Windows 98. If, for example, if you will google the
alt.windows98 newsgroup, for example, you'll see that I've said the same
thing myself *many* times. I am not arguing against that at all.

 

[Stan Brown]

Tue, 14 Feb 2006 20:01:01 -0700 from Ken Blake, MVP

My point is not that one should use any multiple of
RAM. It's that in the absence of any other information on which to base a
change to the Windows default setting, it's best to leave that default
alone, rather than run the risk of making it worse than it already is.

And we agree a thousand percent on that point.

On the other, I'm probably arguing semantics and I doubt it's worth
the effort. The main point, as you and I have both said, is that the
general run of users should not alter Windows XP's default sizing of
the page file.