check SYS(1104) out at:
http//support.microsoft.com/support/kb/articles/Q269/2/84.asp

It also is a documented function in VFP7. Interestingly, it doesn't appear you'll ever need to use it in VFP7. VFP7 (beta) reclaims the memory in all tests *immediately*, without the need to force it.


>Ed,
>While sys(1104) appears to work, what does it do? I can't find it documented anywhere (Help, MSDN, Technet).
>
>>Very interesting. I place the sys(1104) in my form manager class
>>in between each form change, and system memory stays in check nicely on T/S.
>>For VFP5, it seems the _Screen.Lockscreen=.F. or creating a
>>form then destroying it are good tools for T/S applications..
>>
>>>Here's one you don't see everyday.

>>>x = space(16777184)   && VFP memory jumps to between 36-39 MB
>>>_Screen.LockScreen = .F.  && VFP memory drops to 18 MB
>>>release x  && VFP memory drops to normal pre-x levels

>>>It seems VFP puts a lot into system resources and the LockScreen causes these to be re-evaluated.
>>>
>>>HTH.
>>>
>>>>I'm hoping someone knows how to force VFP to do garbage collection.
>>>>For example, on an NT machine with task manager watching the VFP memory allocated, I type:
>>>>
>>>>x = space(16777184) ' VFP allocates between 20-36meg of RAM
>>>>x = Null
>>>>Release x ' RAM stays allocated
>>>>
>>>>CLOSE ALL
>>>>CLEAR ALL ' No luck - still allocated
>>>>
>>>>Is there any way to force memory allocation? SYS(3050) has
>>>>no effect on this kind of memory allocation.
>>>>
>>>>We have a citrix box with many VFP apps running, and if one
>>>>user runs a VFP routine which needs RAM, the RAM is staying
>>>>allocated to that user indefinitely, slowing the other users
>>>>down.
>>>>
>>>>If anyone has any insites, I would be very thankful.
>>>>
>>>>TIA,
>>>>Ed