>I'm not sure I buy that reasoning. If it was a VFP problem, it would exhibit that problem regardless of type of disk. It seems from Ed Rauh's testing that SCSI at least (and Fiber Channel and FireWire [don't remember if he tested those]) don't have this problem.
>

It looks like drivers that respect the disabling of write caching don't have the problem. VSS as well as WinZip saves things correctly round tripping SCT files.

>As for the problem "only" manifesting itself in VFP - all versions of the Fox have had a long and colourful history of pointing out weaknesses in hardware, drivers, and OS. To maximize performance, VFP may be using obscure calls that aren't made in other vendors' products. Or, could it be (*gasp*) that VFP uses undocumented calls that other vendors won't even know about?