Thanks for pointing out the difference of hibernation vs. sleep after a timeout or with keyboards having a special sleep (moon) key. I think the beta version problem also was about that normal sleep state and not only the hibernation "deep" sleep mode. So if that's still not fully fixed, the best idea still is to not leave VFP open, even in coffee breaks. The last hard problem of similar nature I had was about media change events I reacted to with bindevents. It did work in general, but explicitly failed on the notebooks a customer had already ordered in high numbers. We had to skip that feature to automatically react to media change events and instead use a timer polling DRIVETPYE().
I just bring this here, as you also might only have that problem in conjunction with your specific hardware. Which again just means avoid the problem. Turn up sleep times and put some things like hdd drives to a mode never going to sleep, which could help in case it's about behind the scenes used C++ filehandles for workareas becoming invalid. Not handles you have with FOPEN, but handles the VFP runtime has through USE or anything like that.
Bye, Olaf.