>In my situation I am now focusing on what I consider to be an abnormal number of page faults that occur during communications. I noticed that if I send packets across my system I get a predictable number of page faults each and every time. Every time I see a page fault I see the memory usage go up by four bytes for the process I am monitoring and never go down. The memory remains allocated until I quit vfp...
>
>If you haven't done it yet, go to GOOGLE and search for "Process Esplorer". Process Explorer is one of a suite of excelent diagnostic utilities writen by the guys at SysInternals and hosted on the Microsoft site. It allows you to monitor a specific process and not just the whole machine like Task Manager. Try it you'll like it.

Steve,

I do know ProcMon and ProcExp<g>. But for your memory tests: if you exhausted all easy internal vfp tricks (sys1104, sys3050, wait wind timeout) you might try to tweak the memory from the windows angle. Before going to WinAPI, watch memory when the app looses the focus (window size the same) and in contrast if you minimize the window. The latter should trim the WorkingsetSize similar to Api calls (and is a double-edged blade)

HTH

thomas