On 11/2/07, Mat Marcus <mat-lists@emarcus.org> wrote:
My approach is different. You don't have to change the interfaces to use optional_poly (so it is non intrusive with existing hierarchies), and you don't need all this boilerplate code, but you have to use the -> operator to access members of the class.
[snip]
You correctly note that the adobe::poly library requires more machinery to be provided by certain stakeholers. It does not, however, intrude into existing concrete classes.
[snip] Ok, we are talking of two different things: intrusiveness w.r.t. existing hierarchies vs. intrusiveness w.r.t. existing concrete classes. Your example shows that adobe::poly<> is not intrusive w.r.t. existing concrete classes, because it simulates concept maps with the boilerplate code, but you are starting from a class that is not polymorphic (I would do a similar think in those cases, look at my Callable example; the only difference is that the boilerplate code seems to me more simple and natural, because we create a hierarchy of classes, as we would usually do in OOP). I'm concerned, though, of the case in which you already have a hierarchy of polymorphic classes, and want to take advantage of value semantics, and other possible optimizations e.g. small object. If you look at my example on how to rewrite boost::any to have the small object optimization using optional_poly, you will see how an existing hierarchy: class placeholder { public: // structors virtual ~placeholder() { } public: // queries virtual const std::type_info & type() const = 0; }; template<typename ValueType> class holder : public placeholder { public: // structors holder(const ValueType & value) : held(value) { } public: // queries virtual const std::type_info & type() const { return typeid(ValueType); } public: // representation ValueType held; }; is completely reused without changes (I only removed an unused method for clarity of the example). I think this is out of adobe::poly scope, or at least it is never covered in the examples, so I cannot figure out how this would be done in practice. Corrado -- __________________________________________________________________________ dott. Corrado Zoccolo mailto:zoccolo@di.unipi.it PhD - Department of Computer Science - University of Pisa, Italy --------------------------------------------------------------------------