Hi the list,
I know for having read it somewhere that, as does gcc, GPC stops evaluating as soon as it knows what is the result. For example, evaluating for A and B, with A set to 0, it does not evaluate B for it knows it does not worth doing it.
I don't know if this is due to the fact that, at a certain scale, GPC is mixed up with gcc, and here is my question:
Is it a particularity of gcc (injected in GPC) or is it something true for all compilers. At least, can one code with this feature in the back of the mind without fear of borland unportability.
One may object that it does not change anything but efficiency. I have conversely some code that will work if only the first part is evaluated but will bug if the second is (that is quite curious. Don't worry. It works with GPC, will it too with borland?)
Cheers.