What is the design rationale behind the resize method of std::vector?

Question

Many methods within the template class vector take a const reference to value_type objects, for instance:

void push_back (const value_type& val);

while resize takes its value_type parameter by value:

void resize (size_type n, value_type val = value_type());

As a non-expert C++ programmer I can only think of disadvantages with this choice (for instance if size_of(value_type) is big enough stack overflow may occur). What I would like to ask to people with more insight on the language is thus:

What is the design rationale behind this choice?

Answer 1

void resize( size_type count, T value = T() );

This function has been removed from C++11.

C++11 has two overloads of resize():

void resize( size_type count );
void resize( size_type count, const value_type& value);

which is pretty much straightforward to understand. The first one uses default constructed objects of type value_type to fill vector when resizing, the second takes a value from which it makes copies when resizing.