>When i start my application and go in a program how prints two reports, after the firts report is printed and when vfp starts building the second report i became the following error C00000FD
>
>does someone have seen this error before
>
This sounds suspiciously like a problem related to an ill-behaved printer driver; there's been an ongoing problem with drivers from certain manufacturers, notably HP, which do not handle the floating point/MMX registers of the CPU in a consistent and predictable fashion. The symptoms are a wide range of errors immediately following actions that invoke the printer driver, in many cases at the start of a subsequent call.
There's a common way to prove out that this is the problem, and several lines of attack to work around the problem. The simplest thing to do is to try using a known stable, compatible printer driver from a Microsoft operating system distribution rather than the current vendor-supplied driver. In the case of HP particularly, the problems are usually cleared up immediately. Selecting a driver for the test is fairly simple in the case of an HP laser; for testing purposes, almost every laser printer other than PostScript-only lasers can act like an HP laserJet Series II, and the driver is about as simple and stable as can be found. For testing purposes, install the LaserJet Series II printer driver from the operating system CD under whatever version of Windows is being used, and try running your report again. If the problem goes away, you've pinpointed the cause.
What to do about it depends on what version of VFP you're using. If you have VFP6, upgrading to the current version by applying Viwual Studio Service Pack 3 fixes the problem neatly in the vast majority of the cases; MS added code to deal with resetting the procesor after calling the print driver. You'd need to update any systems that have the VFP6 runtime with the updated runtime components from a revised Setup Wizard distribution, and apply SP3 to all systems that have the development version.
On all versions of VFP, you can select an MS-provided driver for a compatible printer that may give up a few model-specific features, bells and whistles in the vendor's supplied drivers. This is the approach I've recommended to many of my clients who have newer HP printers with pretty HP drivers that crash things. As an example, I've had good success with the MS-distributed HP 4 Plus driver in place of the driver for the HP 6 and HP 4xxx series printers; aside from some driver-specific settings that my applications (and frankly, that very few apps other than some page layout software) do not exploit at all, the drivers work just fine, are rock-solid, and are well-behaved.
There's a workaround in the MSKB that suggests making an API _fpreset() call after each invocation of the print driver by VFP; this cleans up the state of the floating point registers in the CPU. it works most of the time, but not always, and it's code that you have to remember to add anywhere that the print driver gets invoked. And there are lots of things that may invoke the print driver even without sending something to the printer. If you can't use an alternative driver, work in a version of VFP earlier than 6, and are willing to make sure that you add the code everywhere it might be needed, it's worth a shot. I've found that it's usually better to switch to a different driver so that my code stays the same, and if the vendor corrects their ill-behaved driver at some point, you can always put the updated driver in place without affecting the application.
>
>Thanks
