RAM

On Wed, 21 Jul 2004 17:45:31 GMT, Ron Martell
wrote:

"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote:


So what you are saying is that I am not freeing up memory
when I defrag and free up the memory? Funny how after I
defrag and/or scrub my memory, but it and my Taskbar clock
TclockEx report the same amount of RAM. So it must work.

You have totally misunderstood the basic concepts of memory management
in Windows. Free memory is more appropriately described as *useless*
memory because that is what it actually represents - memory for which
Windows has so far been totally unable to find any potentially
beneficial use for.



***************

Frequently Asked Questions

Here we answer some frequently asked questions from MemTurbo
users. If you don't see the answer to your question here,
please take a look at the more comprehensive online version
of the FAQ.




What does MemTurbo do?
MemTurbo manages system memory, the paging file, and the way
in which virtual memory is treated in order to maximize
performance. It recovers RAM not currently needed by the
operating system and applications, and recovers memory
leaked by applications. It also can temporarily flush
unused DLLs and libraries out to disk to make room for your
big applications and games (these come back transparently
when needed).

And when they are needed it takes up to 1,000 times longer for Windows
to reload these items from the hard drive than it would to access them
if they had remained in RAM. That is the price of useless////free
memory - slows things down tremendously.


On Windows9X, it also adjusts the system
caching in order to better target the way you use your
particular machine. This can improve disk performance,
gaming performance, and can prevent buffer under-runs that
interfere with, for example, burning CDs.


That is pure balderdash, with a touch of hogwash and a bit of malarkey
thrown in. Note that they do not provide any timed benchmark reports
to substantiate their claims. The windows disk cache requires no
tweaks, except on systems with more than 512 mb of RAM, and that tweak
is a one time entry that takes perhaps 15 seconds to do.



I'm not a Technical person... save the geek-speak and just
tell me how to use it effectively!
Sure! Let it start from your Startup group so that it is
running at all times. After exiting a memory-hungry program
or before launching a new one, press the hotkey (CTRL-ALT-M)
to recover and defragment your RAM. In a matter of seconds
your system should have that "just booted" feel!


RAM fragmentation is another snake-oil myth. RAM is always
fragmented, because Windows puts things where it wants to and there is
no way to control or prevent this.


So plain old defragging is useless and another bit of
bloatware ?


Furthermore RAM fragmentation has
zero repeat zero impact on performance under any circumstances. All
addresses in RAM are equally accessible within the RAM chip access
time, and there is no difference in the time required to switch from
an address at the beginning of RAM to one at the upper end compared to
that required to switch between two adjacent addresses. None. Zero.

How do I stop a RAM recovery in progress?
When a RAM recovery is in progress, as indicated by the
on-screen status display, just press escape. If a
background recovery (triggered by your timer or a memory
alarm level) is running, clicking on the MemTurbo tray icon
will abort the recovery.

MemTurbo should be permanently and totally aborted. :-)


Sometimes my system seems slower after recovering RAM...
why?
If you have your "Target" level of RAM to recover set too
high, MemTurbo may reclaim memory from the file cache, or
flush system DLLs (such as the shell and OLE) out of memory.
When you flip to an application that needs these, they must
be paged back in. Try setting your Target level to a lower
level. Note that for games that do not use much of the
operating system, a higher Target level is better, since
more memory will be available to the game. Remember that
too much free RAM is as good as RAM in your desk drawer:
nice to own but unused! A good compromise takes a while to
discover, but that's why we give you the control to adjust
it!


There is no need to compromise. Period. Just let Windows handle the
memory management and forget about these snake-oil products.


Can MemTurbo make my system unstable?
No. If anything, because it increases the amount of memory
available to applications, your system should become more
stable. It installs no VxDs or drivers and does not modify
your system files in any way.


It will make your system slower in use, and that is enough to condemn
it to the dust bin.

And available RAM has zero repeat zero effect on overall stability.



Does MemTurbo compress memory?
Absolutely not; there is no performance-robbing compression
at all. It simply causes what physical RAM you have to be
used in a more efficient manner, and allows you to reclaim
that memory from applications and the operating system when
you need it most.


More effecient? Slowing down the computer is more efficient?

Why does MemTurbo not always recover up to the Target
setting?
Because if you set the target too high, you've set an
impossible goal! MemTurbo will recover as much RAM as
possible, and can usually get up to your target level
(though the higher the level, the more work MemTurbo must
do, and hence the longer it will take). Note that on
Windows NT, the kernel manages the memory in such a way that
as soon as memory is recovered, it is used by waiting
applications and the operating system, so the displayed
value may never equal the target level. It's still doing
its job, though, even if not apparent!


The target should be to get rid of MemTurbo.


Why can "Program X" seemingly recover more RAM than
MemTurbo?
MemTurbo tries to be pragmatic about its memory recovery.
It doesn't just arbitrarily toss things out of memory to
meet your goal if that means dumping code and data you
really should keep in memory, like parts of the kernel and
so forth. While it would be possible to recover more memory
at times (and early betas of MemTurbo did so), it proved
detrimental to performance, so we strove to keep the
heuristics on how to recover more conservative.


Why do I get "Low Virtual Memory" pop-ups under Windows NT?
Your page file size is too small for proper performance.
Take the system's advice, and increase your paging file size
in your system properties.


Especially because MemTurbo is forcing the system to move a lot of
stuff unnecessarily to the page file so as to increase the supply of
*useless* RAM.



So what you're saying is that if all my RAM is used up I can
run MSWorks 7.0 ? Hate to be the bearer of bad news, but
such is not the case.


Can MemTurbo really recover memory leaks?
While it cannot recover them from the address space (the
application would fault if it did indeed try to use that
memory at some point) it can recover the physical RAM
leaked, making it available for use by the operating system
and other applications.

All it does is force active memory content out to the page file,
thereby slowing the system down because that memory content needs to
be paged back in again when it is needed.



What if it isn't needed? Or if free RAM is "ueless" then is
not also a waste of money and space to have RAM chips
installed in the first place? After all if you have no free
memory, that is tantamount to no memory and thus just let
the Swap file handle the memory. But that will not work as
the system needs RAM in addition to the Swap file.




What is the Memory Load Index?
This is a statistic provided by the system that serves as a
general measurement of how much demand for RAM there
currently is in your system, with 0 being very little and
100 being very much. If you find your system pegged to 100%
shortly after startup, you should seriously consider adding
more physical memory to your system (even MemTurbo won't be
enough by itself in this case).


Can MemTurbo improve gaming performance?
Yes. Let's say you start a game, and part way in, it
suddenly demands memory for graphics, etc. Without
MemTurbo, your system would visibly pause as other
applications and parts of the operating system were
page-faulted out to make room. If you "Defragment and
Recover" before starting the game, MemTurbo should make a
noticeable improvement.


How does MemTurbo prevent buffer underruns while burning
CDs?
In the Windows9X registered version only, it adjusts the
system file caching to ensure that enough cache is set aside
so that the data you are burning to a CD is available,
rather than having to fight with the CD on the I/O bus for
the next block of data.

***************

For a system that is bloatware and a resource hog, from what
I gather XP is moreso than 98/98SE, why isn't such a good
thing?

Because MemTurbo is based on a totally false premise regarding the
value of *useless* RAM.


Also why does XP require a faster processor and more RAM
than 98/98SE ? At best, unless I upgrade my MOBO, I can
only get WindowsME.

Because it is a bigger operating system with improved capabilities
over Windows Me. It is based on the Windows NT kernel rather than the
Windows 9x kernel used in Windows 95/98/Me, and the NT based versions
of Windows have always been larger and more demanding than the 9x
based versions.

Windows XP is designed to use the higher speed CPUs, larger memory
modules, and larger hard drives that were generally available in 2001,
and many of the features and functions included in Windows XP require
this higher performance and increased capacity.

My first copy of Norton Utilities would fit on a single 360K 5.25 inch
diskette. Today it requries several hundred megabytes of space on a
CDROM. It is called evolution, and we are still only half-way through
inventing the PC.


You'll have to do better than mentioning that such a piece
of software is bad. That's like saying fire is bad. It is
bad if uncontrolled, it is also required in some cases as in
a forest fire to release the seeds from the pine cones.
Without fire, there would be many a cold meal.


Fire has many redeeming features. MemTurbo has none.

Good luck


Ron Martell Duncan B.C. Canada


Sorry, but I'm still not convinced as I have not seen any
conclusive proof one way or the other. All system
environments are different more than likely, so what may be
true for one individual may not be true for another. Having
a Compaq it has a certain thinking as to the way it wants to
behave, luckily having flashed the BIOS removed that
thinking, it also removed the capablity to use my Quick
restore CD, at least the drivers are available.

--
Sincerely, | (©) (©)
| ------ooo--(_)--ooo------
Andrew H. Carter | /// \\\
d(-_-)b |

That's correct. You are only wasting RAM when you are running utilities
that prevent Windows from using it.
--
Jeff Richards
MS MVP (DTS)
"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote in message
...
snip
But I often, depending on my operations, get down to 1MB
(actually a directory comparison the other day brought it
down to 60KB). So the RAM isn't wasted, rather money well
spent.

On Thu, 22 Jul 2004 19:24:35 +1000, "Jeff Richards"
wrote:

That's correct. You are only wasting RAM when you are running utilities
that prevent Windows from using it.

RAM isn't my problem, it is system resources which when I
close some proggies it's as though I haven't closed it for
the percentage remains the same. So I'm thinking some TSRs
are at play.

When I booted my computer earlier todayI had System
Resources at 86%, now they are at 67%, no matter what I do,
I won't be able to get back down or up to 86% +/- until I
reboot. Sure I may gain some back, but still!

--
Sincerely, | (©) (©)
| ------ooo--(_)--ooo------
Andrew H. Carter | /// \\\
d(-_-)b |

I rarely had resources in excess of 50% when I was running W98 and my
computer worked fine. A low level of resources only becomes a problem when
you approach 0% or run out of them. You cannot increase resources--they are
fixed chunks (64 kb) of memory. What point are you trying to make?
--
Regards

Ron Badour, MS MVP for W98
Tips: http://home.satx.rr.com/badour
Knowledge Base Info:
http://support.microsoft.com/default.aspx?pr=kbinfo

"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote in message
...
On Thu, 22 Jul 2004 19:24:35 +1000, "Jeff Richards"
wrote:

That's correct. You are only wasting RAM when you are running utilities
that prevent Windows from using it.

RAM isn't my problem, it is system resources which when I
close some proggies it's as though I haven't closed it for
the percentage remains the same. So I'm thinking some TSRs
are at play.

When I booted my computer earlier todayI had System
Resources at 86%, now they are at 67%, no matter what I do,
I won't be able to get back down or up to 86% +/- until I
reboot. Sure I may gain some back, but still!

--
Sincerely, | (©) (©)
| ------ooo--(_)--ooo------
Andrew H. Carter | /// \\\
d(-_-)b |

"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote:


RAM fragmentation is another snake-oil myth. RAM is always
fragmented, because Windows puts things where it wants to and there is
no way to control or prevent this.


So plain old defragging is useless and another bit of
bloatware ?


Defragging of a disk drive is not useless, no. And I never said it
was.

But RAM defragging is totally useless, and companies that promote this
are merely proving oncd again that they have zero understanding of how
RAM works or how Windows manages it.

snip


Especially because MemTurbo is forcing the system to move a lot of
stuff unnecessarily to the page file so as to increase the supply of
*useless* RAM.



So what you're saying is that if all my RAM is used up I can
run MSWorks 7.0 ? Hate to be the bearer of bad news, but
such is not the case.



There is, or should be, absolutely no reason why you cannot run MS
Works 7.0 (or any other installed application) even if the unused RAM
is at or near zero when you launch the application.

What exactly happens when you try to do this? Is there some sort of
an error message produced? The only circumstance where I can imagine
this happening is if MemTurbo (or some other miguided piece of
crudware) has placed a maximum size limit on the virtual memory swap
file. That will produce "out of memory" problems and is always a bad
idea in Windows 95/98/Me.



All it does is force active memory content out to the page file,
thereby slowing the system down because that memory content needs to
be paged back in again when it is needed.



What if it isn't needed? Or if free RAM is "ueless" then is
not also a waste of money and space to have RAM chips
installed in the first place? After all if you have no free
memory, that is tantamount to no memory and thus just let
the Swap file handle the memory. But that will not work as
the system needs RAM in addition to the Swap file.


Having RAM in excess of what can be used by Windows is a waste, and
some people do so. I see systems quite regularly where the user has
512 mb of RAM with Windows 98 and, because of the number and type of
applications that they use, they would not suffer any loss in
performance if they reduced the RAM to 256 mb and there would not be
much of hit if they dropped back to 128 mb. Same thing with Windows
XP, only bigger numbers. Some people are going for 4 gb of RAM on
Windows XP systems for home use, usually on the concept that "if some
is good then more is better" and they get no improvement in
performance over what they would have with only 1 gb of RAM or perhaps
even with only 512 mb.

If the total memory load placed on the computer were known and
predictable then you could configure with exactly that much RAM and
performance would still be optimal. But memory requirements are
dynamic, and users are not 100% consistent in the programs and data
files that they use.

Free memory is not tantamount to no memory. Just try booting a
computer with no RAM installed. :-).

The basic fact is that Windows will always attempt to find some use,
anything whatever that might potentially be of some benefit, rather
than just leaving the RAM sitting there idly going to rot and doing
absolutely no good for anybody. And just as soon as some better use
comes along for some of that RAM then Windows will instantaneously
drop the more trivial usages so as to free up whatever is now
required.

So Windows will at times create a huge disk cache, using as much as
80% of the installed RAM if there are no other requirements for that
RAM. It will also retain in RAM the program code from recently closed
applications, in anticipation of that application being launched again
in the near future.

What MemTurbo and similar products due is to create a fake need for
large amounts of RAM, forcing Windows to reduce the disk cache and/or
to drop recently closed programs from RAM to meet this new RAM
requirement. Then MemTurbo cancels the RAM request, resulting in a
quantity of unused RAM; and also slower performance because of the
smaller disk cache and especially slower startup times should a
recently closed application be launched again.

Here is an experiment you can try.

Boot up the computer.
Launch a large application program such as Microsoft Word and time how
long it takes to open.
Then close it and launch it again. Time the second startup as well -
it should be very significantly faster.
Now launch MemTurbo and have it free up as much RAM as it possibly
can.
Now launch Microsoft Word again. This time it will take as long, or
almost as long, as it did the first time.
That is a brief demonstration of how MemTurbo hurts the overall
performance of a computer.



snip

Sorry, but I'm still not convinced as I have not seen any
conclusive proof one way or the other. All system
environments are different more than likely, so what may be
true for one individual may not be true for another. Having
a Compaq it has a certain thinking as to the way it wants to
behave, luckily having flashed the BIOS removed that
thinking, it also removed the capablity to use my Quick
restore CD, at least the drivers are available.

It is your right to not be convinced. I could be a bit facetious and
suggest that you might also find a membership in the flat earth
society appealing as well, but I will save that for another time and
place, like over a cold beer or two. :-)

Good luck


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

"The reason computer chips are so small is computers don't eat much."

On Thu, 22 Jul 2004 08:44:11 -0500, "Ron Badour"
wrote:

I rarely had resources in excess of 50% when I was running W98 and my
computer worked fine. A low level of resources only becomes a problem when
you approach 0% or run out of them. You cannot increase resources--they are
fixed chunks (64 kb) of memory. What point are you trying to make?

The more resources one has, the "faster" the comp will run,
so is there a way to defrag the system resources. BUt
apparently there isn't.

--
Sincerely, | (©) (©)
| ------ooo--(_)--ooo------
Andrew H. Carter | /// \\\
d(-_-)b |

On Thu, 22 Jul 2004 21:04:14 GMT, Ron Martell
wrote:

"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote:


RAM fragmentation is another snake-oil myth. RAM is always
fragmented, because Windows puts things where it wants to and there is
no way to control or prevent this.


So plain old defragging is useless and another bit of
bloatware ?


Defragging of a disk drive is not useless, no. And I never said it
was.

But RAM defragging is totally useless, and companies that promote this
are merely proving oncd again that they have zero understanding of how
RAM works or how Windows manages it.

snip


Especially because MemTurbo is forcing the system to move a lot of
stuff unnecessarily to the page file so as to increase the supply of
*useless* RAM.



So what you're saying is that if all my RAM is used up I can
run MSWorks 7.0 ? Hate to be the bearer of bad news, but
such is not the case.



There is, or should be, absolutely no reason why you cannot run MS
Works 7.0 (or any other installed application) even if the unused RAM
is at or near zero when you launch the application.

What exactly happens when you try to do this? Is there some sort of
an error message produced? The only circumstance where I can imagine
this happening is if MemTurbo (or some other miguided piece of
crudware) has placed a maximum size limit on the virtual memory swap
file. That will produce "out of memory" problems and is always a bad
idea in Windows 95/98/Me.



As to recreating, it may take a while, but I suppose I could
max out all resources and RAM, then try to recreate it.

I let Windows manage the swap and in System Properties
Performance File System Troubleshooting, all the boxes are
unchecked.

Running Tasks:

Kernel32.dll
MSGSRV32.EXE
Spool32.exe
Mprexe.exe
Avgserv9.exe
MMTASK.TSK
Explorer.exe
Taskmon.exe
Systray.exe
Type32.exe
Em_exec.exe
Memturbo.exe
Wmiexe.exe
Ddhelp.exe
Avgcc32.exe
Smc.exe
Rnaapp.exe
Tapisrv.exe
Agent.exe
Editpad.exe
Cliptray.exe
Msinfo32.exe


Startup:

MemTurbo Startup Group
TClockEx Registry (Per-User Run)
ScanRegistry Registry (Machine Run)
TaskMonitor Registry (Machine Run)
SystemTray Registry (Machine Run)
IntelliType Registry (Machine Run)
EM_EXEC Registry (Machine Run)
LoadPowerProfile Registry (Machine Run)
AVG_CC Registry (Machine Run)
TimeSink Ad Client Registry (Machine Run)

**TimeSink: This is part of my WHOIS app. I could disable
it, but is on my low end of snooping concerns. The proggie
will function without it, but the interface is bland without
it, I'm peculiar that way.


Tweak UI Registry (Machine Run)
LoadPowerProfile Registry (Machine Service)
Avgserv9.exe Registry (Machine Service)



All it does is force active memory content out to the page file,
thereby slowing the system down because that memory content needs to
be paged back in again when it is needed.



What if it isn't needed? Or if free RAM is "ueless" then is
not also a waste of money and space to have RAM chips
installed in the first place? After all if you have no free
memory, that is tantamount to no memory and thus just let
the Swap file handle the memory. But that will not work as
the system needs RAM in addition to the Swap file.


Having RAM in excess of what can be used by Windows is a waste, and
some people do so. I see systems quite regularly where the user has
512 mb of RAM with Windows 98 and, because of the number and type of
applications that they use, they would not suffer any loss in
performance if they reduced the RAM to 256 mb and there would not be
much of hit if they dropped back to 128 mb. Same thing with Windows
XP, only bigger numbers. Some people are going for 4 gb of RAM on
Windows XP systems for home use, usually on the concept that "if some
is good then more is better" and they get no improvement in
performance over what they would have with only 1 gb of RAM or perhaps
even with only 512 mb.

If the total memory load placed on the computer were known and
predictable then you could configure with exactly that much RAM and
performance would still be optimal. But memory requirements are
dynamic, and users are not 100% consistent in the programs and data
files that they use.

Free memory is not tantamount to no memory. Just try booting a
computer with no RAM installed. :-).

The basic fact is that Windows will always attempt to find some use,
anything whatever that might potentially be of some benefit, rather
than just leaving the RAM sitting there idly going to rot and doing
absolutely no good for anybody. And just as soon as some better use
comes along for some of that RAM then Windows will instantaneously
drop the more trivial usages so as to free up whatever is now
required.

So Windows will at times create a huge disk cache, using as much as
80% of the installed RAM if there are no other requirements for that
RAM. It will also retain in RAM the program code from recently closed
applications, in anticipation of that application being launched again
in the near future.

What MemTurbo and similar products due is to create a fake need for
large amounts of RAM, forcing Windows to reduce the disk cache and/or
to drop recently closed programs from RAM to meet this new RAM
requirement. Then MemTurbo cancels the RAM request, resulting in a
quantity of unused RAM; and also slower performance because of the
smaller disk cache and especially slower startup times should a
recently closed application be launched again.

Here is an experiment you can try.

Boot up the computer.
Launch a large application program such as Microsoft Word and time how
long it takes to open.
Then close it and launch it again. Time the second startup as well -
it should be very significantly faster.
Now launch MemTurbo and have it free up as much RAM as it possibly
can.
Now launch Microsoft Word again. This time it will take as long, or
almost as long, as it did the first time.
That is a brief demonstration of how MemTurbo hurts the overall
performance of a computer.



I'll try that. But I use MemTurbo in Manual mode meaning
that I control when to defrag, having it run automatically
does slow things down as I have almost nil use of the PC
until it is done (manual or automatic)



snip

Sorry, but I'm still not convinced as I have not seen any
conclusive proof one way or the other. All system
environments are different more than likely, so what may be
true for one individual may not be true for another. Having
a Compaq it has a certain thinking as to the way it wants to
behave, luckily having flashed the BIOS removed that
thinking, it also removed the capablity to use my Quick
restore CD, at least the drivers are available.

It is your right to not be convinced. I could be a bit facetious and
suggest that you might also find a membership in the flat earth
society appealing as well, but I will save that for another time and
place, like over a cold beer or two. :-)



Sounds good. Domestic or Imported? German or Egyptian?
Funny, one thinks of Germany being the capital of beer and
the source of it, but the Egyptians were the first makers of
beer, though such was by accident.

Their hard bread, still having yeast would be dipped in
water to soften it, the bread would leave behind small
amounts of yeast and flour, all the ingredients for the
"lifeblood" of the male human.

--
Sincerely, | (©) (©)
| ------ooo--(_)--ooo------
Andrew H. Carter | /// \\\
d(-_-)b |

"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote in message
...
snip
The more resources one has, the "faster" the comp will run,

You cannot increase the amount of resources, so perhaps you are referring to
free resources. That's simply not so. The computer runs at exactly the same
speed regardless of the level of resource usage, until the level gets so low
that applications cannot claim the resources they need. If a program you are
trying to start has a large claim on resources then this may start to occur
when the level has got down to below about 10%. Apart from the possibility
that this might happen, preventing you from running a program, the amount of
free resources has no effect on the system at all.

so is there a way to defrag the system resources.

There is no such thing as fragmentation as far as resources are concerned,
and therefore no need to defrag resources.
--
Jeff Richards
MS MVP (DTS)

If Works requires a private copy of a DLL, and if MemTurbo is able to unload
the default copy of the DLL before Works 7 is started, then it would appear
that 'freeing' RAM enables Works 7 to run. This appearance would be
reinforced by the fact that, in the above scenario, running Works 7
immediately after startup would also work (because the default DLL never got
loaded).

This is, of course, an error in the installation of Works and has nothing to
do with freeing or defragging RAM.
--
Jeff Richards
MS MVP (DTS)
"Ron Martell" wrote in message
...
"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote:


RAM fragmentation is another snake-oil myth. RAM is always
fragmented, because Windows puts things where it wants to and there is
no way to control or prevent this.


So plain old defragging is useless and another bit of
bloatware ?


Defragging of a disk drive is not useless, no. And I never said it
was.

But RAM defragging is totally useless, and companies that promote this
are merely proving oncd again that they have zero understanding of how
RAM works or how Windows manages it.

snip


Especially because MemTurbo is forcing the system to move a lot of
stuff unnecessarily to the page file so as to increase the supply of
*useless* RAM.



So what you're saying is that if all my RAM is used up I can
run MSWorks 7.0 ? Hate to be the bearer of bad news, but
such is not the case.



There is, or should be, absolutely no reason why you cannot run MS
Works 7.0 (or any other installed application) even if the unused RAM
is at or near zero when you launch the application.

What exactly happens when you try to do this? Is there some sort of
an error message produced? The only circumstance where I can imagine
this happening is if MemTurbo (or some other miguided piece of
crudware) has placed a maximum size limit on the virtual memory swap
file. That will produce "out of memory" problems and is always a bad
idea in Windows 95/98/Me.



All it does is force active memory content out to the page file,
thereby slowing the system down because that memory content needs to
be paged back in again when it is needed.



What if it isn't needed? Or if free RAM is "ueless" then is
not also a waste of money and space to have RAM chips
installed in the first place? After all if you have no free
memory, that is tantamount to no memory and thus just let
the Swap file handle the memory. But that will not work as
the system needs RAM in addition to the Swap file.


Having RAM in excess of what can be used by Windows is a waste, and
some people do so. I see systems quite regularly where the user has
512 mb of RAM with Windows 98 and, because of the number and type of
applications that they use, they would not suffer any loss in
performance if they reduced the RAM to 256 mb and there would not be
much of hit if they dropped back to 128 mb. Same thing with Windows
XP, only bigger numbers. Some people are going for 4 gb of RAM on
Windows XP systems for home use, usually on the concept that "if some
is good then more is better" and they get no improvement in
performance over what they would have with only 1 gb of RAM or perhaps
even with only 512 mb.

If the total memory load placed on the computer were known and
predictable then you could configure with exactly that much RAM and
performance would still be optimal. But memory requirements are
dynamic, and users are not 100% consistent in the programs and data
files that they use.

Free memory is not tantamount to no memory. Just try booting a
computer with no RAM installed. :-).

The basic fact is that Windows will always attempt to find some use,
anything whatever that might potentially be of some benefit, rather
than just leaving the RAM sitting there idly going to rot and doing
absolutely no good for anybody. And just as soon as some better use
comes along for some of that RAM then Windows will instantaneously
drop the more trivial usages so as to free up whatever is now
required.

So Windows will at times create a huge disk cache, using as much as
80% of the installed RAM if there are no other requirements for that
RAM. It will also retain in RAM the program code from recently closed
applications, in anticipation of that application being launched again
in the near future.

What MemTurbo and similar products due is to create a fake need for
large amounts of RAM, forcing Windows to reduce the disk cache and/or
to drop recently closed programs from RAM to meet this new RAM
requirement. Then MemTurbo cancels the RAM request, resulting in a
quantity of unused RAM; and also slower performance because of the
smaller disk cache and especially slower startup times should a
recently closed application be launched again.

Here is an experiment you can try.

Boot up the computer.
Launch a large application program such as Microsoft Word and time how
long it takes to open.
Then close it and launch it again. Time the second startup as well -
it should be very significantly faster.
Now launch MemTurbo and have it free up as much RAM as it possibly
can.
Now launch Microsoft Word again. This time it will take as long, or
almost as long, as it did the first time.
That is a brief demonstration of how MemTurbo hurts the overall
performance of a computer.



snip

Sorry, but I'm still not convinced as I have not seen any
conclusive proof one way or the other. All system
environments are different more than likely, so what may be
true for one individual may not be true for another. Having
a Compaq it has a certain thinking as to the way it wants to
behave, luckily having flashed the BIOS removed that
thinking, it also removed the capablity to use my Quick
restore CD, at least the drivers are available.

It is your right to not be convinced. I could be a bit facetious and
suggest that you might also find a membership in the flat earth
society appealing as well, but I will save that for another time and
place, like over a cold beer or two. :-)

Good luck


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

"The reason computer chips are so small is computers don't eat much."

It was pool league last night hence the late response. I agree with what
Jeff said--it makes no difference at all until you approach zero (annoying
notices) and once you get to zero.

Defrag resources? It is only a 64 kb chunk of memory so just how fragmented
could it get? And since memory works so blazingly fast, I think there is no
speeding it up at least from a fragmentation standpoint.

If you cut back on the needless programs started at boot, the system will
*boot* a lot faster and you will have more resources available to start with
which might preclude running out later in a computing session.

--
Regards

Ron Badour, MS MVP for W98
Tips: http://home.satx.rr.com/badour
Knowledge Base Info:
http://support.microsoft.com/default.aspx?pr=kbinfo

"Andrew H. Carter (Applied ROT 17 Left, for Email do 17 Right)"
wrote in message
...
On Thu, 22 Jul 2004 08:44:11 -0500, "Ron Badour"
wrote:

I rarely had resources in excess of 50% when I was running W98 and my
computer worked fine. A low level of resources only becomes a problem
when
you approach 0% or run out of them. You cannot increase resources--they
are
fixed chunks (64 kb) of memory. What point are you trying to make?

The more resources one has, the "faster" the comp will run,
so is there a way to defrag the system resources. BUt
apparently there isn't.

--
Sincerely, | (©) (©)
| ------ooo--(_)--ooo------
Andrew H. Carter | /// \\\
d(-_-)b |