Can we have RPN back?

[orcinus]

(03-28-2014 03:34 PM)John R. Graham Wrote:  [*]Behavior of stack lift. Starting with an empty stack, "2 [Enter] x" results in 4 on legacy RPN calculators and an error on RPL calculators (and the Prime).

Nope. Enter behaves as DUP when there's nothing on the entry line.
On RPL calculators (confirmed on a 48 and two 49s) and Prime (at least in the emulator). That's how it always was, AFAICR.

(03-28-2014 03:34 PM)John R. Graham Wrote:  [*]Auto-copy of T. As operators consume the contents of the stack, on legacy RPN calculators, the contents of T is duplicated into Z so that the stack never truly empties. (In fact, the stack always contains 4 numbers.) On RPL machines, the stack can actually empty, resulting in operation not being able to complete for lack of arguments.

I don't think i understand this one (bear with me, my history with HP begun with RPL, not RPN). Are you saying that on the classic RPN models, as you continue hitting DROP, the contents of T fills out the stack levels beneath it? Isn't that an accident waiting to happen?