EInvalidOP when Filtering

[DaveH]

We have been chasing down an intermittent error "Invalid Floating Point Operation" for over a year. Unfortunately it can not be predicted when the error will occur.

Using a component named ExceptionMagic, however, we have been able to trace the error to the "ApplyFilterTo" function inside OracleData.pas.

But now we have another problem... there does not seem to be any use of floating point operations in that function. Can you help in any way?

Dave

 

[Marco Kalter]

The ApplyFilterTo function calls your OnFilterRecord event handler. It subsequently calls TExpression.AsBoolean. Both functions can use floating point operations, but ExceptionMagic should have reported that in its stack dump. This is not the case?

------------------
Marco Kalter
Allround Automations