Thanks for bringing this to our attention.

I can reproduce the problem you describe. I can reproduce it when requery() is preceeded by =, or without the = (but intermittently in this case). I don't see the problem when the return value is stored to a variable. I also found that AError() will populate an error array after the failed requery with error 1526, "Connectivity error: Unable to retrieve specific error information. Driver is probably out of resources". This makes it possible to trap for the error condition.

In cases where the requery fails, if I conditionally reexecute it it seems to always succeed the 2nd time.

I changed your code in the procoption method to demonstrate:

=REQUERY("ORDER") 

* Added these lines
=AError(atest)
if atest[1] = 1526
   =REQUERY("ORDER") 
endif

* reset error handler to error other than 1526
on error *
error 1
on error
* End of added lines

IF RECCOUNT("ORDER") < 1 OR ( order.orderid <> nOrderID )

I have passed this information up the line.

Thanks!

>Several days ago I posted about a serious bug in VFP8 that basically renders the VFP8 based applications unreliable. I didn't get very many responses, probably because many here thought it was something "I was doing." Well, here ya grow! Looks like the problem is more than just REQUERY()! Kind of surprised that Microsoft would not have chimed in here on this one. Oh well!
>
>I tried to post the code here but it wouldn't allow me to post it.
>
>Pickup up the programs at:
>ftp:\\www.interactivedatasystems.com\rebug.zip.
>
>Two programs here, Northwind.prg to generate a test DBC and rebug.prg to that contains the code necessary to show that REQUERY() has issues and/or VFP8 is unstable. You will need access to SQL 2000 and the classes from the FFC directory under HOME().
>
>Rebug.prg
>*-- Demonstration of a Major REQUERY() bug in VFP v8.
>*--
>*-- The following class libraries are from the
>*-- VFP8 FFC directory. Notice that if you move
>*-- the object creation up, say below _BASE, the problem
>*-- may or may not go away. You may also need to add
>*-- more class libraries to reproduce. On our development
>*-- machines, this is all it takes. Go ahead, try it and then
>*-- move the object creation up a few classes. As you approach the
>*-- the top, the problem goes away voila!
>*-- Demonstration of a Major REQUERY() bug in VFP v8.
>*--
>*-- The following class libraries are from the
>*-- VFP8 FFC directory. Notice that if you move
>*-- the object creation up, say below _BASE, the problem
>*-- may or may not go away. You may also need to add
>*-- more class libraries to reproduce. On our development
>*-- machines, this is all it takes.
>
>Northwind.prg
>*-- Generate the required views.
>
>Enjoy!