>PMI, Ed, but that's not an entirely accurate statement. The metadata files are opened and used internally by VFP a bit differently that applications use tables with memo fields and are, therefore, less likely to corrupt.
John, Ed, and others,
Just to clear the issue up, the thing that affects the likelihood of any file being damaged is the "window of vulnerability". That window the duration that the file is being written to. Because of how VFP stored memo data, an update usually takes longer than the same update to a dbf file. The longer it takes to update the larger the window of vulnerability and according to statistics the higher the incidence of damage.
As John suggested VFP writes to the metadata tables internally, meaning there is no interpretation going on the execution is faster, thus reducing the window of vulnerability and the likelihood of damage.
I have quite a few systems out there using memo fields and do not have problems with FPT fiel damage. OTOH, there are systems out there that do have FPT corruption problems. I can only surmize that I must be doing something differently than those systems do it.