I’l wait for Bob’s take on this, but I think it’s a kind of half-empty, half-full situation. I don’t think there’s any real reason to do it one way or the other. It just makes sense in my mind not to claim success until you’ve ruled out failure.

The code uses html tags (<pre> </pre>), none of the square bracket tags, because I really don’t like how any of those plugins work. Until I get four hundred comments a day, it’s not too burdensome to fix it myself. Those of us with OCD have lots of energy for that sort of thing.

You didn’t spot that I missed the Function keyword in the Function signature though

You’re right, I should start showing ByVal and ByRef. In fact, I should probably start using them in my own code. I’ll start tomorrow.

As I mentioned in DDoE, I have gotten into the habit of rarely passing a function result back as a function result, I rather pass all results back (I often require more than one, even if only a non-critical return code in addition to the functional value) as ByRef parameters. My function then passes back whether it was a critical failure or not. For example

Public Function MyFunction( _
	ByVal input As String, _
	ByRef output1 As Long, _
	ByRef output2 As String) As Boolean

	MyFunction = True
	On Error Goto MyFunction_err

	'some code that does stuff and sets up output1 and output2

MyFunction_exit:
	Exit Function

MyFunction_err:
	MyFunction = False
	Resume MyFunction_exit
End Function

 
and call it like so

	If Not MyFunction("Bob", mpAge, mpSex) then

		'handle this error
		Exit Function
	End IF

	'now process the mpAge and mpSex variables returned

I like the boolean approach best, especially for the straightforward routines that either should work or completely fail. Then just test for that in your calling procedure.