OldxBaser
Programmer
- Nov 5, 2009
- 24
We have noticed an increased report of Error 41 (missing or corrupt FPT) over the last half year from our client base.
Most of our users run our very large application in a shared network environment, on various combinations of platforms. Some have dozens of users logged in simultaneously with no problems; file/record locking works consistently well.
However, this slight increase in reports of a 'corrupt' or 'missing' FPT seems to have coincided with the increased attention to Windows security, especially with Vista and with various AVS software.
I don't know the real sequence of file-opening when VFP opens a DBF that has a CDX and FPT. Is it possible that the DBF opens first, and then it detects the need to open the associated FPT and CDX before allowing access to the DBF data? If so, what if the resident real-time AVS is still scanning the FPT when a VFP command tries to open a file in exclusive mode? Would this not generate an error 41?
We have had some occurrences of client data sent to us that fails with this error on the client network but runs perfectly here, and we are trying to find a generic resolve to this.
Any comments on this?
Most of our users run our very large application in a shared network environment, on various combinations of platforms. Some have dozens of users logged in simultaneously with no problems; file/record locking works consistently well.
However, this slight increase in reports of a 'corrupt' or 'missing' FPT seems to have coincided with the increased attention to Windows security, especially with Vista and with various AVS software.
I don't know the real sequence of file-opening when VFP opens a DBF that has a CDX and FPT. Is it possible that the DBF opens first, and then it detects the need to open the associated FPT and CDX before allowing access to the DBF data? If so, what if the resident real-time AVS is still scanning the FPT when a VFP command tries to open a file in exclusive mode? Would this not generate an error 41?
We have had some occurrences of client data sent to us that fails with this error on the client network but runs perfectly here, and we are trying to find a generic resolve to this.
Any comments on this?