I use numeric return codes from functions quite a bit and it helps with debugging and testing. This is very helpful for complex functions that have many exit points.

I use values > 0 for true and the value may be different to indicate which case the function returned a true value. The same for return codes that I interpret as .f. I have a couple of logging functions that I can turn on at design time or at runtime to show parameters passed and return codes to verify that the function is executing the way that I thought it would.

If MYFUNCTION(parameterlist...) > 0
*!* returned .t.
else
*!* returned .f.
endif