Question

I am reading Wikipedia article on difference btween JAVA and C++. One difference is that C++ offers ‘multiple binary compatibility standards’. Could you explain what this means, or hint at a good reference. I have a clue that it means that binary ‘written with’ C++ is very portable, can be used on any OS or environment. I would like to have confirmation and more precision. What is it all about?

How to generate binaries? What make it not portable?

Thanks and regards.

Answer 1

What does int mean, exactly? When calling a function with 2 parameters – do you put the first one first on a stack or last; or do you have a structure on a heap and point to it? Do you allow an unknown number of arguments being passed into a function? How do you treat strings; arrays? Do you allocate on stack, heap, or in a private memory block? Do you mangle function names (to allow for overloads) or use them as input in source code? Do you align members in a structure on 8 bit, 16 bit, or 32 bit boundary?

All those questions (and many more) are making a great deal of difference to how one binary calls another and the answers are not as simple most of the time.

Java doesn’t offer much in terms of how exactly a binary layout is being done (as it’s a VM, after all) while C++ offers great flexibility to accommodate almost any imaginable requirement out there – thus it “offers binary compatibility standards”, unlike Java (in your example)