Some claim how good a job Win XP is doing managing memory. I have a different take on it - and recent experiments with the FSX demo only added to my suspicions.I played the Fritz 9 chess program which allows you to manually set its internal hastable size (so it can store intermediate results). I set it to around 600 MB. I have 1 GB of RAM. After finishing the game of chess I completely exited the Fritz - I made sure it was no longer running. Then I proceeded to run the FSX demo - it never fully loaded - I got some sort of error and the demo exited. I suspected the problem might be not enough memory - after I rebooted my PC I had no trouble running the demo. My suspicion is that XP for some reason hogs the memory from the program that no longer is running. My virtual memory settings: initial size 720 MB, maximum size 1534 MB.Michael J.

Remember that applications, and not just the OS, can cause memory problems. It used to be more of a problem in the past, but some apps will not properly free their memory after they are closed. Buggy beta apps like the FSX demo are particularly prone to such behavior...so I am not surprised you see it. I've seen it in FS9 with improperly coded gauges, too. And of course we've all seen it with the FS9 default.xml file bug.RE: your virtual mem settings, some would say that it's best from a performance standpoint to set your swap file min and max sizes to the same.Rhett

>It used to be more of a problem in the past,>but some apps will not properly free their memory after they>are closed. This is big news to me and if true I guess this must be unique to windows environment issue. Under Unix no matter how the application misbehave - the operating system will always do the final "clean up" after the application is no longer running. Michael J.

Have you repeated this or was this a one time event?Early this week, I finished rendering a 10 minute video in Vegas and then loaded up FSX and it worked fine. It may have been a fluke you experienced, or something other than a memory management issue. I will admit, the memory reallocation is handled differently and more than likely less efficiently in Windows than Unix, but in response it is put more in the hands of the programmers to handle.

Windows only releases the memory it has acquired itself. Applications must themselves release memory they have acquired. Using C++, memory is acquired by the new operator and released by delete operator. These operators create and delete objects in C++.The reason for this (I believe, but am not certain) is to allow for applications sharing memory/objects in a multi-tasking environment. For example, one application might create objects than another application could use. If they were automatically deleted when the first application closed, then the second application could crash.I'm not familar enough with UNIX to know how it deals with this situation.

Mybe Fritz never released the memory properly... What was you available physical before/during/after running Fritz?I have an external system monitor www.pertelian.com and trust me, XP does release memory correctly, even after a 10 hour flight with FS...-Paul

>Have you repeated this or was this a one time event?You may be right, it might have been a fluke. I tried it today and it worked. But to really repeat the same thing I did last time I have to play a really long game of chess (like 2+ hrs) and see what happens then. Fritz may acquire more and more memory as the game gets longer. I will test it again tonight.Michael J.http://www.precisionmanuals.com/images/for...argo_hauler.gifhttp://sales.hifisim.com/pub-download/asv6-banner-beta.jpg

While it is a good style to free the memory when a program is terminated, it is not needed. The operating system frees *all* memory and resources when a program is ended. Even the resources a program "forgot" to free.Alex