In the development version we have redesigned the header so that for most words nt=xt (i.e., NAME>INTERPRET does nothing for these words).
We are now considering whether to use the PFA (body address) rather than the CFA as xt/nt. Now would be a good time for such a change, because the new header has already caused code that relies on a specific header layout to break. Advantage: >BODY and BODY> become noops, and you do not need to use them. Disadvantages: 1) Because >BODY and BODY> become noops, and you don't notice it if you forget to use them, portability to other Forth systems (where >BODY currently is usually not a noop) suffers. 2) If you are using @ rather than >CODE-ADDRESS ( xt -- addr ) to get at the code address, your code would break from such a change. Anything else? Do you have code that would break from such a change? - anton
