Re: [DotGNU]random question

[CH Gowri Kumar]

> Hmm... right now it makes lots of sense to use C because the rest of the
> framework is in C :)
Ofcourse.

But why was C used in  instead of C++, when we are
trying to simulate the same through user level programming.

To be more precise by what I mean simulating, this is what I have
observed(may be wrong)
In C++
Class class{
        //...
        Data members
        //....
        Function Members
}

In C:
typedef struct class{
        //..
        Data Members
        //...
        Function Pointers (to the function members)
}class;

And then we have an initalize function which does the assigning of the
function pointers to the appropriate functions.
class c;
Initialize(&c);

After that it to call member functions:
c->MemberFunction(&c,..)...


> Every single bit of OO you add increases size , complexity and speed
> penalties to the system. This idea is almost identical to OO , but
> without inheritance or constructors or base classes. Which is to say
> "virtually no overhead" :).
This is the reason hear from a lot of people.(I have also quoted the same
in the question which i posed in the firstmail).

But just wanted to know what is the increase in overhead between the
implementations of C and C++.

By way of simulating we are incurring the pointer indirection everytime.
Also all the structure elements have function pointers increasing the size.
And debugging becomes **difficult** with function pointers.

Just curious to know the reason :)

Regards,
Gowri Kumar