Stack left dirty in case of error

[camillew]

trafficstars

The stack is left with some intermediary data in case of error:

8 0 / -> 1 [and an error] 8 0 MOD -> +NaN [and an error]

It would be great to have a behavior similar to the RPL calculators where the arguments were restored in case of error.

[c3d]

This behavior will "appear" when I get to implementing the undo stack and LastX.

Until then, you are correct that the state of the stack in case of error is not guaranteed.