>>Ed,
>>
>>>>you're poking at cockroaches with a blunt stick<<
>>
>>That is how it feels. Thanks very much for the additional information.
>
>I hope that you didn't take the 'additional information' as absolving VFP from virtually any C000005.
>
>VFP6 SP3 fixed many C00005 problems, obviously all within VFP. They may even have all been fixed by a single revision, but they did appear under various conditions.
>
>From what I've seen reported here, and a couple of experiences of my own, I think it is still fair ball to have VFP itself as a prime suspect in many of the occurrences.
>
>Who knows - maybe the SP will address some of them.
>

First, we're discussing VFP7, not 6, and the memory manager issue that caused the vast majority of C5 errors was addressed in VFP6 SP3 as you've stated. Second, I really believe that installation and distribution cause many of the reported errors, where components are not upgraded properly. My experience with VFP7 is that the InstallShield Ltd that ships with VFP7 does not adequately ensure that all dependencies throughout the station environment are addressed, especially for the 'copy some stuff and run' crowd. I suggest that cruising over to http://www.installsite.com would be an educational experience for most VFP developers who aren't immersed in Windows internals. I certainly found the web site to be well worth investigating, first using VSI, and later, the current IS Pro for Windows. I also recommend that upgrading to full IS Express 3 is well worth the money.

For me, VFP7 is delivering reliable executables and COM servers in production environments, some of which are operating in complex and critical roles. I'm using Win2K SP2 and XP Pro to run things, and require the systems to have more than a minimum memory configuration (256MB-512MB is generally comfortable, and PC133 SDRAM is cheap. RAMBUS is still a bit pricey, and from my POV, is not improving system performance enough for most apps.)

>Cheers