>>Yeah, but if Lmargin happens to have a value close to Imargin, you couldn't tell.
>>
>>The order will control the value at the time of evaluation. It might be using the value from the line above (which, if I remember correctly was close to the line where it was right.)
>
>It's an "i". I double-checked. The variables are declared in the order that I listed them. I don't understand why it would be using a value from the line above? I can't see a situation where you would want it to start using values from a different group on random lines.

The order you showed should work fine. Can you verify that sum, count, and so forth are NOT checked?