Re: static initialization order fiasco

From:

"Alf P. Steinbach" <alfps@start.no>

Newsgroups:

comp.lang.c++

Date:

Wed, 30 Jul 2008 10:39:33 +0200

Message-ID:

<prWdnXNDFMnctA3VnZ2dnUVZ_rHinZ2d@posted.comnet>

* anon:

When I was checking this:
http://www.parashift.com/c++-faq-lite/ctors.html#faq-10.12
(questions until 10.16)

This code:

//////////////////////////////////////////////////
#include <iostream>
#include <memory>

class a
{
    public:
        a() { std::cout<<"a()"<<std::endl; }

        ~a() { std::cout<<"~a()"<<std::endl; }
};
class b
{
    public:
        b() { std::cout<<"b()"<<std::endl; }

        ~b() { std::cout<<"~b()"<<std::endl; }
};
class c
{
    public:
        static a& k()
        {
            static std::auto_ptr< a > p( new a );
            return *p;
        }
        static b& j()
        {
            static b* p( new b );
            return *p;
        }
};
int main()
{
    c::k();
    c::j();
}
//////////////////////////////////////////////////

the output is:
a()
b()
~a()

So, the question is why the faq doesn't suggest this solution (with
auto_ptr) to the static initialization order fiasco?

The FAQ doesn't suggest using auto_ptr because for ordinary sized objects it
doesn't help with anything:

static a& k() { static a theObject; return theObject; }

That's called a "Meyers singleton".

Use of auto_ptr is an orthogonal issue, addressing the problem of allocating
Very Large objects.

See also the following FAQ items.

Cheers & hth.,

- Alf

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?