Swap a and b’s values – but do every read and write through pa and pb, never by writing a or b
directly. You’ll still need one ordinary temporary variable to hold a value mid-swap, exactly like a swap without
pointers would; the only thing that changes is that every access to a and b themselves goes through a pointer
instead.
This is worth getting comfortable with by hand here, in main, because the very next lesson wraps this identical
logic inside a function – at which point it stops being optional and becomes the only way to write a function that
can swap its caller’s variables at all.