>As I understand the article, it would be better to define your own "types", depending on how the variable will be used.
>
>For example, applying this to VFP, if you have a TextBox, derived from a class to handle monetary amounts (currency type; appropriate InputMask), instead of calling it TxtPurchasePrice, you might name it MoneyPurchasePrice, for example. The same (I think) if you create the TextBoxes directly, without using classes. This would give you much more useful information than the mere fact that it is a TextBox.

I use the txt,cmb,edt,lbl kinds of prefixes - and unless the textbox is associated with a DBF field - I use long descriptive names.

I typically use long descriptive function and variable names. But some recent work with JSP and Java has influenced me. For example,I would use a flag name with with an "ll" prefix (local logical):
llIsTheFileWellFormed=GET_WellFormedStatus()
But I find myself dropping the 'll' convention:
isTheFileWellFormed=GET_WellFormedStatus()

Wondering if this was what the author meant?

I have noticed that C++ and Java developers have some fairly 'obvious' and natural conventions. (I used the pronoun "I" 6 times - appolgies for the first person laziness:-)