[boost] Re:an xml-binary persistence library

From: Robert Ramey (ramey_at_[hidden])
Date: 2003-04-21 11:14:04

Vladimir Prus wrote:

>That's interesting. I wonder if 'const-correctness', here, is just a
>theoretical question, or can cause real problems. For example (using
>describe-based scheme):

>   class C {
 >  public:
 >       C() ...
 >       template<class D>
        // "const" here advertises that invocation of the function won't change the state
        // of the instance.  This isn't true for loading.
 >       void describe(D& d) const {
 >               d & m_i;
 >       }
        // My assumption has been that "const int m_i" can only be at object creation time.
        // Situations such as this motivated me to permit non-default constructors
        // submitted system.  Now I wonder.  In any case I prefer to minimize const casting
        // at least at the user level.
 >       const int m_i;
>   };

with alternate save load the save can be const while the load isn't.  Thats the best 
from a const- point of few.

>a.store_polymorphic_pointer_with_known_type<Update_event>(&ev);

This is supported in the reviewed version.  The requirement for forward declaration
of derived classes will be lifted.  I believe that would address your case.

Robert Ramey